Cannabis and Your Medications

Key Takeaways

• Online interaction checkers list several hundred drugs for cannabis and cannabidiol. The large majority of those entries trace back to laboratory experiments on liver enzymes that have never been tested in patients.
• Five interactions have actual human evidence behind them: clobazam, valproate, transplant anti-rejection drugs, warfarin, and the additive drowsiness that comes from combining cannabis with alcohol or sedatives.
• Dose is what decides whether any of this applies to you. The studies that found real interactions used 700 to 2,900 mg of cannabidiol a day, far above what most retail products deliver in a serving.
• The transplant signal is the exception, appearing at 100 to 300 mg a day, which is within reach of ordinary products.
• Taking cannabidiol with a high-fat meal raised blood levels roughly fourfold versus an empty stomach, a larger swing than most interactions on any warning list.

---

Search for cannabis drug interactions and the numbers arrive fast. Widely used interaction checkers list several hundred medications alongside cannabis and cannabidiol. Read a few articles and you will find lists of drugs to avoid, usually with no indication that some entries rest on decades of patient data and others rest on a single experiment in a dish.

That distinction is the whole story. Cannabinoids do slow down a family of liver enzymes called cytochrome P450, and those enzymes process a large share of prescription drugs. Demonstrating that in a laboratory is straightforward. Demonstrating that it happens in a living person, at the amount that person actually takes, is a different problem, and for most of the drugs on those lists nobody has done it. A small number of interactions have been measured in patients, some of them repeatedly, and those are the ones that should change what you do.

Where the Big Numbers Come From

A laboratory can show that cannabidiol slows a particular enzyme. What it cannot show is whether enough cannabidiol ever reaches your liver to matter.

Researchers have tried to close that gap with modeling, feeding measured laboratory values into calculations that predict what would happen at real-world doses. Those predictions separate sharply by product and by route. A 700 mg oral dose of cannabidiol was predicted to raise levels of the sedative midazolam nearly fifteenfold. The same modeling predicted a much smaller effect from tetrahydrocannabinol, and smaller again when it was inhaled instead of swallowed, because inhaling bypasses the first pass through the gut and liver where most of this interference happens [1].

Two things follow. The first is that "cannabis" is not one exposure. A high-dose oral cannabidiol product and an inhaled flower product are different pharmacological situations, and a warning that lumps them together is telling you almost nothing. The second is that these are predictions awaiting confirmation in patients. The researchers say so themselves. A predicted fifteenfold change is a reason to study something, not a finding about you.

The Five That Are Not Theoretical

Clobazam, a seizure medication. This is the best-documented interaction in the field. In 13 children taking both drugs, blood levels of clobazam's active breakdown product rose by an average of 500 percent over four weeks, and 10 of the 13 developed side effects that eased once their clobazam dose was lowered [2]. The mechanism is simple enough. Cannabidiol slows the enzyme that clears that breakdown product, so it builds up. Anyone on clobazam should have both the drug and its metabolite measured when cannabidiol is started or the dose changes.

Valproate, another seizure medication. The concern here is liver enzymes, not drug levels. In a study of 39 adults and 42 children taking cannabidiol alongside various seizure drugs, with doses escalated from 5 up to 50 mg per kilogram daily, liver enzyme readings ran significantly higher in the participants also taking valproate [3]. Valproate can affect the liver on its own, so the combination is what matters. Liver testing before starting, and again during any dose increase, is the standard response.

Transplant anti-rejection drugs. A transplant patient taking 2,000 to 2,900 mg of cannabidiol daily showed roughly a threefold rise in tacrolimus levels [4]. More striking, a small series of seven kidney transplant recipients using only 100 to 300 mg a day for chronic pain still saw tacrolimus levels move enough that two of them needed dose reductions, while the patients on cyclosporine stayed steady [5]. This is the one place where ordinary product doses have produced a documented signal. The consequence of getting it wrong is losing a transplanted organ, so a trough level drawn within a week or two of any change is a reasonable thing to ask for.

Warfarin, a blood thinner. A 44-year-old man with a mechanical heart valve, whose clotting numbers had held between 2.0 and 2.6 for at least six months, watched them climb as his cannabidiol dose was raised every two weeks. His warfarin dose was eventually cut by about 30 percent, and he had no bleeding complications [6]. That stable six-month baseline is what makes a single case persuasive, since it rules out most other explanations for the drift. The evidence here comes from case reports, not from trials, so how often this happens is unknown, but an INR check within a week or two of any change in cannabis use is simple and already routine.

Alcohol and sedatives. This one is not about liver enzymes at all. In a driving simulator study, 18 participants used vaporized cannabis with and without a low dose of alcohol. Cannabis alone, at a blood THC level around 8 micrograms per liter, impaired lane control about as much as a 0.05 breath alcohol reading, and combining the two produced roughly the sum of both effects [7]. Nothing exotic is happening. Two things that make you drowsy make you drowsier together. The participants were young and healthy, so for an older adult already on sleep or anxiety medication, the real fall risk is probably larger than this study shows.

Dose Is What Decides Whether This Applies to You

The studies above did not use products from a dispensary shelf. They used 700 mg, 750 mg, and in the transplant case up to 2,900 mg of cannabidiol daily. Retail products are typically labeled in tens of milligrams per serving. That gap is the single most useful thing to understand, and it is missing from nearly every warning list you will find.

So the first question is not which drugs interact. It is how many milligrams you are actually taking. Check the label, multiply by servings, and compare that number to the hundreds used in the research. For readers who have not thought about their own dose in milligrams before, Answering Questions About Cannabis Microdosing covers starting small. If you are an order of magnitude below, the metabolic concerns in most of those lists have little to do with you. The transplant finding is the exception that proves the rule, since it appeared at 100 to 300 mg a day, and anyone on anti-rejection medication should treat any amount as worth disclosing.

Two other things separate a careful attempt from a reckless one. Take it the same way each day, because consistency matters more than timing. And tell your prescriber before you start, not after something looks strange, since the useful version of that conversation happens while there is still a baseline to compare against.

One thing to avoid outright: do not stop a prescribed medication because an interaction checker flagged it. The flag usually reflects a laboratory finding, the checker does not know your dose, and stopping a seizure drug or a blood thinner carries a certain risk in exchange for a theoretical one.

Food Changes How Much Gets In

In a controlled trial, people took the same 750 mg dose of cannabidiol fasted and after meals. A high-fat meal raised total blood exposure 3.8-fold and peak concentration 5.2-fold compared with fasting. A low-fat meal and whole milk produced smaller but still substantial increases [8].

Read that against the warning lists. A four- to fivefold swing from breakfast is larger than most of the interactions those lists are warning you about. It also means that a dose taken with dinner one day and on an empty stomach the next is effectively two different doses, which makes any effect harder to interpret, good or bad. Picking one routine and staying with it is the closest thing to dose control available without a prescription.

Cannabidiol Does Not Soften THC

A common claim in cannabis retail is that cannabidiol takes the edge off tetrahydrocannabinol, so a balanced product should feel gentler than a THC-dominant one at the same THC dose. A randomized trial tested exactly that. Eighteen adults took brownies containing either a THC-dominant extract or a cannabidiol-dominant extract, both delivering 20 mg of THC.

The cannabidiol-dominant version produced more anxiety, more sedation, more memory difficulty, a higher heart rate, and worse performance on thinking and coordination tests [9]. Blood levels of THC were also higher, which is the likely explanation: cannabidiol appears to slow the breakdown of THC, leaving more of it circulating.

This is a single small trial using edibles, so how far it extends to other products and ratios is unknown. But the direction is the opposite of the claim. For oral products at these ratios, adding cannabidiol did not blunt the effects of THC. It amplified them.

The Bottom Line

The long interaction lists are not fabricated, but they are answering a laboratory question when yours is a clinical one. For most people taking retail-strength products, the metabolic concerns behind those lists have not been shown to matter. For people on clobazam, valproate, transplant medication, or warfarin, and for anyone combining cannabis with alcohol or sedatives, the evidence is real and the response is monitoring, not avoidance.

Knowing your milligrams, keeping your routine steady, and telling your prescriber before you start covers most of what you can do without a clinic visit. What none of it supports is stopping a prescribed medication on the strength of a warning generated by a database.

For the underlying studies, the evidence grading behind these conclusions, and the drug classes where the research has genuinely come up empty, see the companion research article, Cannabinoid Drug Interactions and Polypharmacy: Grading the Evidence From Confirmed to Contradicted.

Frequently Asked Questions

Does CBD interact with more medications than THC? In laboratory testing and modeling, yes, mainly because the cannabidiol doses used in medicine are so much larger. A 700 mg oral cannabidiol dose and a typical amount of inhaled tetrahydrocannabinol are not comparable exposures [1].

Is smoking or vaping safer than edibles for interactions? Modeling predicts a smaller metabolic effect from inhaling, because inhaled cannabinoids skip the first pass through the gut and liver where most interference occurs [1]. No study has compared the two routes head to head for interaction outcomes.

My checker says my medication has a major interaction. Should I stop taking it? No. Bring the result to your prescriber or pharmacist instead. These flags generally reflect laboratory findings and do not account for your dose, and stopping a prescribed medication carries its own risk.

Does CBD damage the liver? The liver signal in the research is tied specifically to taking cannabidiol alongside valproate, not to cannabidiol on its own [3]. If you take valproate, liver testing is the appropriate step.

I take a statin. Should I be worried? Statins appear on interaction lists on the strength of laboratory data only. No clinical study has tested cannabis or cannabidiol against statins in patients, so the honest answer is that this remains an open question rather than a known risk.

---

References

• [^1]: Bansal S, Paine MF, Unadkat JD. Comprehensive Predictions of Cytochrome P450 (P450)-Mediated In Vivo Cannabinoid-Drug Interactions Based on Reversible and Time-Dependent P450 Inhibition in Human Liver Microsomes. Drug Metab Dispos. 2022;50(4):351-360. https://pubmed.ncbi.nlm.nih.gov/35115300/
• [^2]: Geffrey AL, Pollack SF, Bruno PL, Thiele EA. Drug-drug interaction between clobazam and cannabidiol in children with refractory epilepsy. Epilepsia. 2015;56(8):1246-1251. https://pubmed.ncbi.nlm.nih.gov/26114620/
• [^3]: Gaston TE, Bebin EM, Cutter GR, Liu Y, Szaflarski JP. Interactions between cannabidiol and commonly used antiepileptic drugs. Epilepsia. 2017;58(9):1586-1592. https://pubmed.ncbi.nlm.nih.gov/28782097/
• [^4]: Leino AD, Emoto C, Fukuda T, Privitera M, Vinks AA, Alloway RR. Evidence of a clinically significant drug-drug interaction between cannabidiol and tacrolimus. Am J Transplant. 2019;19(10):2944-2948. https://pubmed.ncbi.nlm.nih.gov/31012522/
• [^5]: Cuñetti L, Manzo L, Peyraube R, Arnaiz J, Curi L, Orihuela S. Chronic Pain Treatment With Cannabidiol in Kidney Transplant Patients in Uruguay. Transplant Proc. 2018;50(2):461-464. https://pubmed.ncbi.nlm.nih.gov/29579828/
• [^6]: Grayson L, Vines B, Nichol K, Szaflarski JP. An interaction between warfarin and cannabidiol, a case report. Epilepsy Behav Case Rep. 2017;9:10-11. https://pubmed.ncbi.nlm.nih.gov/29387536/
• [^7]: Hartman RL, Brown TL, Milavetz G, Spurgin A, Pierce RS, Gorelick DA, Gaffney G, Huestis MA. Cannabis effects on driving lateral control with and without alcohol. Drug Alcohol Depend. 2015;154:25-37. https://pubmed.ncbi.nlm.nih.gov/26144593/
• [^8]: Crockett J, Critchley D, Tayo B, Berwaerts J, Morrison G. A phase 1, randomized, pharmacokinetic trial of the effect of different meal compositions, whole milk, and alcohol on cannabidiol exposure and safety in healthy subjects. Epilepsia. 2020;61(2):267-277. https://pubmed.ncbi.nlm.nih.gov/32012251/
• [^9]: Zamarripa CA, Spindle TR, Surujunarain R, Weerts EM, Bansal S, Unadkat JD, Paine MF, Vandrey R. Assessment of Orally Administered Δ9-Tetrahydrocannabinol When Coadministered With Cannabidiol on Δ9-Tetrahydrocannabinol Pharmacokinetics and Pharmacodynamics in Healthy Adults: A Randomized Clinical Trial. JAMA Netw Open. 2023;6(2):e2254752. https://pubmed.ncbi.nlm.nih.gov/36780161/