Cannabis and Medication Interactions: What Labels and Clinical Studies Establish

The clearest human evidence concerns prescription cannabidiol, particularly the approved oral cannabidiol formulation covered by its drug label.

By Cannible Editorial

2026-08-10

The clearest human evidence concerns prescription cannabidiol, particularly the approved oral cannabidiol formulation covered by its drug label. Demonstrated interactions include increased exposure to clobazam’s active metabolite, increased stiripentol exposure, increased everolimus exposure, increased liver-enzyme elevations with valproate, and increased sedation risk with other central nervous system depressants or alcohol. Evidence for nonprescription cannabis products and many other prescription medicines is less certain and should not be treated as equivalent to the label evidence. 1 2

General information only: This article does not provide individualized medical advice or instructions to change, start, or stop a medication. An adult who uses prescription medicine should discuss cannabis or cannabinoid use, including product and dose information when known, with the prescriber or pharmacist.

How to read the evidence

The strongest information here comes from the official prescribing information for prescription cannabidiol and from a human pharmacokinetic trial in healthy volunteers. The trial examined cannabidiol with clobazam, stiripentol, and valproate; it found a 3.4-fold increase in exposure to N-desmethylclobazam, a smaller increase in stiripentol exposure, and no clinically relevant change in valproate exposure. The study was open-label and conducted in healthy volunteers, so its findings describe measured drug levels and tolerability in that setting rather than every cannabis product or every patient population. 2

Interactions DEMONSTRATED in humans

With prescription cannabidiol, clobazam exposure itself changed little in the clinical trial, but exposure to its active metabolite N-desmethylclobazam increased 3.4-fold. The drug label warns that this may increase the risk of clobazam-related adverse reactions. Stiripentol exposure also increased, while the trial found no clinically relevant effect of cannabidiol on valproate exposure. 1 2

The label reports an approximately 2.5-fold increase in everolimus exposure when everolimus is taken with prescription cannabidiol. It also states that exposure to other orally administered P-gp substrates, including sirolimus, tacrolimus, and digoxin, may increase. These are label-established pharmacokinetic concerns, not proof that every person taking one of these medicines will experience harm. 1

The label states that combining prescription cannabidiol with valproate increases the incidence of liver-enzyme elevations. It also warns that combining cannabidiol with other central nervous system depressants, including alcohol, may increase sedation and somnolence. These warnings establish clinically relevant safety concerns, even though the healthy-volunteer trial found no clinically relevant change in valproate exposure. 1 2

For warfarin, the available human evidence is limited and low quality. A systematic review found seven case reports involving warfarin; six reported increased INR values, with reported maximum changes ranging from 0.4 to 9.61 where a maximum was available. One report found no INR change after four days of medical-cannabis exposure. Each case report concerns a single patient, and these reports do not establish how often an interaction occurs or that it will occur with every cannabis or cannabinoid product. 3

For anticoagulants other than warfarin, the systematic review found no evidence supporting an interaction. That conclusion is about the evidence identified in the review, not a guarantee that an interaction is impossible. 3

Enzyme-level THEORETICAL interactions not shown clinically

Cannabinoids and many prescription medicines use overlapping CYP enzyme pathways. Laboratory and pharmacology evidence describes cannabidiol as an inhibitor of several enzymes, including CYP1A2, CYP2C19, and UGT1A9, and as both an inducer and inhibitor of CYP2B6. The label therefore identifies possible exposure changes for medicines handled by CYP1A2, CYP2B6, CYP2C8, CYP2C19, and UGT1A9. These enzyme-level findings are not, by themselves, clinical proof that a particular patient will have an interaction. 1 4

Examples in the label include possible increased exposure to theophylline and tizanidine through CYP1A2, possible clinically significant interactions with CYP2C8 substrates, and increased concentrations of CYP2C19 substrates. For CYP2C19 medicines whose effect depends mainly on active metabolites, the label describes a different theoretical concern: cannabidiol may reduce active-metabolite concentrations and potentially reduce efficacy. The label says the clinical importance of some of these changes is unknown or depends on the cannabidiol dose. 1

A separate review found that in-vitro cannabinoid effects on CYP enzymes are mostly inhibitory and that cannabidiol is a potent inhibitor of many CYPs, but concluded that the clinical relevance of reported CYP-level interactions remains unclear. Predictions based only on shared enzymes should therefore be kept separate from interactions demonstrated in human studies. 4

Monitoring recommendations

The prescription-cannabidiol label recommends monitoring therapeutic drug levels of everolimus and adjusting its dosage accordingly when cannabidiol is initiated. It recommends considering therapeutic drug monitoring and dosage reduction for other orally administered P-gp substrates, such as sirolimus, tacrolimus, and digoxin. Any medication changes belong with the treating prescriber or pharmacist. 1

The label recommends monitoring for stiripentol-related adverse reactions and considering a clobazam dosage reduction if known clobazam adverse reactions occur with concomitant prescription cannabidiol. It also identifies possible dosage modification for certain CYP2B6, CYP2C8, CYP2C19, CYP1A2, and UGT1A9 substrates when clinically appropriate. 1

For warfarin, the case-report review supports clinical recognition of a potential interaction and the need for a formal interaction study; it does not establish a universal monitoring schedule. The review of narrow-therapeutic-index medicines also found unexpected prescribed-drug serum levels in 18 of 31 reports, reinforcing the value of transparent communication with healthcare professionals about cannabinoid use. 3 5

Symptoms that warrant urgent medical attention

Urgent medical attention is warranted for signs of serious bleeding or major mental-status change. In the warfarin case reports, one patient had bleeding requiring hospitalization, reversal, and treatment after INR elevations; the broader review also identified bleeding risk and altered mental status among reported adverse events. Marked sedation or somnolence is also a clinically important warning when prescription cannabidiol is combined with other central nervous system depressants or alcohol. 3 5 1

Liver-related concerns also deserve prompt medical attention when prescription cannabidiol is used with valproate, because the label reports increased liver-enzyme elevations with that combination. A prescriber or pharmacist should be told about all cannabis or cannabinoid exposure and all prescription medicines; they can interpret symptoms, laboratory results, and drug levels in the individual clinical context. 1 5