What the Liver Does to an Edible Before You Feel Anything

When THC is absorbed from the gut, it travels through the hepatic portal vein and is exposed to liver enzymes before it reaches the systemic circulation. CYP2C9 in particular hydroxylates it at the 11 position, producing 11-hydroxy-THC, a metabolite that is itself active at cannabinoid receptors.

The clearest human demonstration of this is a ratio. In controlled dosing work published in Clinical Pharmacology and Therapeutics, the plasma ratio of 11-hydroxy-THC to THC was roughly 1:10 to 1:20 after intravenous dosing but approached 0.5 to 1:1 after oral dosing. Same drug, same person, entirely different metabolic exposure.

Patients who report a bad experience with cannabis are disproportionately describing an edible. The usual explanation offered to them is that edibles are stronger, which is imprecise. They are slower to declare themselves, longer lasting, and metabolically different, and each of those properties contributes independently to overshoot.

Oral bioavailability of THC is low and variable, in the range of 10 to 20 percent in controlled work. Low and variable is a worse combination than low alone, because it means the same labeled dose of the same product genuinely does land differently in different people.

Wall and colleagues, in Clinical Pharmacology and Therapeutics in 1983, gave THC to men and women both intravenously and orally, and tracked the parent compound and its metabolites. The finding most relevant here is the ratio of 11-hydroxy-THC to THC in plasma: approximately 1:10 to 1:20 after intravenous administration, compared with roughly 0.5 to 1:1 after oral administration.

That difference is not about how much cannabis was taken. It is about where the drug goes first. Inhaled THC crosses the alveolar membrane into pulmonary circulation and reaches the brain before the liver has had much opportunity to act on it. Swallowed THC is absorbed into the portal circulation and passes through hepatocytes on the way to anywhere else.

The same study estimated oral bioavailability at roughly 10 to 20 percent by comparing area under the curve across routes, and reported a terminal half-life for THC of 25 to 36 hours regardless of route or sex. So a swallowed dose delivers less parent THC to the circulation, delivers proportionally far more of an active metabolite, and clears slowly.

The most useful modern dataset on this is from Spindle and colleagues at Johns Hopkins, published in the Journal of Analytical Toxicology in 2020. Seventeen healthy adults, eight of them women, none of whom had used cannabis in at least the previous two months, ate brownies containing 0, 10, 25, or 50 mg of THC across four double-blind outpatient sessions. Blood and oral fluid were sampled at baseline and for eight hours afterward.

Whole blood concentrations of THC and 11-hydroxy-THC peaked 1.5 to 2 hours after eating, declined steadily, and typically returned to baseline by eight hours. Concentrations of 11-nor-9-carboxy-THC and its glucuronide ran higher than either active compound and were frequently still detectable at the eight-hour mark. Women showed higher peak concentrations of THC and of every metabolite measured, which the authors attributed at least partly to lower body weight and body mass index.

The authors also made the route comparison explicit: after oral consumption, THC appeared in blood much later and at far lower peak concentrations than has been observed with inhaled cannabis. That is worth sitting with. Lower blood THC, later, and yet a subjectively stronger and longer experience for many people. Blood THC is a poor proxy for effect after an edible.

This is the claim repeated most often and supported least well. The primary human comparison dates to 1973, when Lemberger and colleagues published in the Journal of Clinical Investigation a study of nine casual cannabis smokers given 1 mg of 11-hydroxy-THC or 1 mg of THC intravenously.

What they observed was a difference in timing. Marked tachycardia and a psychological high occurred within 3 to 5 minutes of 11-hydroxy-THC. After THC, the peak high was delayed by 10 to 20 minutes. Psychological effects correlated well with plasma concentrations of unchanged 11-hydroxy-THC. The authors’ conclusion was that THC is converted to 11-hydroxy-THC in humans and that the metabolite is in part responsible for the psychological effects.

Faster onset at an equal intravenous dose is a real finding. It is not the same as being two or three times more potent, which is the number that circulates online without an attached citation. No modern controlled human study has established a potency ratio between the two compounds, and the existing comparison enrolled nine people more than fifty years ago. The honest statement is that 11-hydroxy-THC is psychoactive, acts quickly once it is in circulation, and is generated in much larger relative amounts after oral dosing.

Genetics contribute measurably. Sachse-Seeboth and colleagues, in Clinical Pharmacology and Therapeutics in 2009, gave oral THC to 43 healthy volunteers genotyped for CYP2C9. Carriers of two copies of the reduced-function CYP2C9*3 allele had a median THC area under the curve three times higher than people with two normal-function copies, and a 70 percent lower area under the curve for 11-nor-9-carboxy-THC. They also showed a trend toward increased sedation.

That is exactly the pattern you would predict if the first-pass hydroxylation step were impaired: more parent drug surviving the liver, less downstream metabolite, and more effect from the same swallowed dose.

Sex and body composition contribute as well, as the Hopkins brownie study showed. So does what else is in the medication list, since CYP2C9 and CYP3A4 handle a long roster of common drugs. A 2025 retrospective chart review in the Journal of Cannabis Research examined 71 oral cannabis users who had undergone pharmacogenomic testing and found atypical variants in CYP2C9 in 31 of them and in CYP2C19 in 37. The authors were careful to frame their work as hypothesis generating, noting the absence of published clinical outcomes tying these variants to patient response. That caution is appropriate and should be carried forward.

The clinical consequence of a slow, variable, metabolically transformed dose is visible in acute care data. Monte and colleagues reviewed five years of cannabis-coded emergency visits at a large urban academic hospital in Colorado, publishing in Annals of Internal Medicine in 2019. Of 2,567 visits judged at least partly attributable to cannabis, 238 (9.3 percent) involved edible products.

Compared with inhaled cannabis, edible-related visits were more likely to involve acute psychiatric symptoms (18.0 percent versus 10.9 percent), intoxication (48 percent versus 28 percent), and cardiovascular symptoms (8.0 percent versus 3.1 percent). The proportionality is the striking part: edibles accounted for 10.7 percent of cannabis-attributable visits between 2014 and 2016 while representing only 0.32 percent of total cannabis sold in Colorado measured in kilograms of THC.

Dose matters, and it is measurable. In a 2026 crossover trial published in JAMA Network Open, 25 healthy adults consumed brownies containing 0, 10, or 25 mg of THC with or without alcohol and completed simulated driving assessments. Every active condition except 10 mg THC alone degraded driving performance. Combining cannabis with alcohol produced greater impairment than either alone, and standard field sobriety testing missed impairment that the driving simulator detected.

The pharmacokinetic core of this is strong. Controlled, blinded, dose-ranging administration on a research unit with mass spectrometric quantification is the right design for the question, and the 2020 Hopkins brownie study did it in cannabis-naive adults, which removes the tolerance confound that muddies work in frequent users.

The route comparison rests on a 1983 study that remains the definitive human mass-balance and ratio work. It is old, and its analytical methods are dated, but no one has repeated it and the finding has been consistent with everything measured since.

The potency claim about 11-hydroxy-THC is the weak link. One small intravenous study from 1973 is the entire human comparative base. It is enough to establish that the metabolite is active and fast. It is not enough to support a numeric potency multiplier.

Blood concentrations are not effects. The Hopkins study measured what was in the blood, not how impaired or how comfortable anyone was. Peak blood concentration at 1.5 to 2 hours does not mean peak subjective effect at 1.5 to 2 hours, and the lag between the two is part of why people redose too early.

Sample sizes are small throughout. Seventeen adults in the brownie study, 43 in the genotype study, nine in the 1973 comparison, 25 in the driving trial. These are the best available studies and they are not population-scale.

The emergency department analysis is retrospective, single-center, and dependent on self-reported route and dose. Its central comparison, edible share of visits versus edible share of the legal market, is a reasonable signal but not a rate. It cannot tell you the risk per person or per dose.

This evidence does not establish that 11-hydroxy-THC is a specific number of times more potent than THC. That figure has no controlled human source behind it.

It does not identify a safe starting dose for any individual. It shows that 10 mg of THC in a brownie did not measurably degrade simulated driving in occasional users while 25 mg did, which is one study, in one population, on one outcome.

It does not show that CYP2C9 genotyping improves cannabis dosing. The association between CYP2C9*3 and higher THC exposure is real, and no trial has tested whether knowing a patient’s genotype in advance changes any outcome.

Cannabis education for patients tends to focus on the amount of THC in a product. The pharmacology says that the amount is only one of three variables, and that route and individual metabolic capacity carry comparable weight. An edible converts a fixed label into a highly variable delivered exposure in a way that inhalation does not.

This also reframes what a dispensary label communicates. Ten milligrams of THC in a gummy is an accurate statement about the product and a poor prediction about the person. The gap between those two is filled by first-pass metabolism, and closing it clinically means starting lower and waiting longer than the label implies is necessary.

Nearly every bad cannabis experience a patient describes to me in detail turns out to be an edible, and nearly every one follows the same script. They took a dose, felt nothing at forty minutes, took more, and then both doses arrived together around the ninety minute mark.

The fix is not complicated and it is not about willpower. It is about knowing that the delay is metabolic and not a sign that the dose was too small. I tell patients to treat an edible as a two-hour commitment before any judgment about whether it worked. If that seems inconvenient, inhalation gives feedback in minutes and is the better tool for anyone who needs to titrate by feel.

The other thing I say plainly: the first time you try a new edible product, the dose that matters is the smallest one you can measure, not the smallest one printed on the package. Those are frequently not the same number.

Oral THC passes through the liver before reaching the brain, and a large share of it is converted to 11-hydroxy-THC, an active metabolite generated in roughly equal proportion to the parent drug after an edible but at a fraction of that proportion after inhalation. Blood concentrations peak at 1.5 to 2 hours, oral bioavailability is only 10 to 20 percent and varies with genotype, sex, and body composition, and the subjective effect lags the measured peak. Start low, wait at least two hours before any redose, and treat edibles as a different drug exposure rather than a different flavor of the same one.

The finding to carry forward is the ratio: oral dosing generates far more 11-hydroxy-THC relative to THC than inhalation does, and that is a measured human result rather than an inference. The finding to leave behind is any specific multiplier describing how much stronger 11-hydroxy-THC is. It is active, it is fast once it reaches the circulation, and beyond that the human data do not support a number.