Botanicals

CBD: One FDA-Approved Use and Thin Evidence for the Rest

CBD has one FDA-approved use — Epidiolex, for rare epilepsies — and much thinner evidence for anxiety, pain and sleep. What the trials show, where the law stands, and why drug interactions matter.

CBD: One FDA-Approved Use and Thin Evidence for the Rest

Key takeaways

  • CBD has exactly one FDA-approved medical use: Epidiolex for certain epilepsy syndromes, at doses far above anything sold over the counter
  • Evidence for sleep and anxiety is promising but not conclusive — mostly small short-term trials; a 12-week placebo-controlled pain trial found no effect at retail-level doses
  • CBD does not get you high; it will not produce intoxication
  • Product labeling is often inaccurate — only 31% of 84 online products tested in a JAMA study were labeled within 10% of their actual CBD content, and 21% contained THC
  • Drug interactions are real — CBD inhibits CYP3A4 enzymes that metabolize many medications, and high doses have caused liver-enzyme elevations in healthy adults

What CBD is

Cannabidiol (CBD) is a non-intoxicating compound from cannabis and hemp - it will not get you high. It is sold as oils, gummies, and topicals for sleep, anxiety, pain, and more.

Chemically it is one of more than a hundred cannabinoids in the plant. Unlike THC, it does not bind strongly to the brain’s CB1 receptor, which is why it does not intoxicate; instead it acts on a spread of other targets - serotonin 5-HT1A receptors, TRPV1 channels, the enzyme that breaks down the body’s own endocannabinoid anandamide - and, importantly, on the liver enzymes that metabolize other drugs. How the endocannabinoid system works is covered separately.

What it is actually proven to do

CBD has exactly one FDA-approved use: a prescription drug, Epidiolex, for two rare, severe childhood epilepsies (Dravet and Lennox-Gastaut syndromes) (FDA). The label was later extended to seizures associated with tuberous sclerosis complex. Everything else - the gummies and tinctures - is sold as an unregulated supplement for uses the FDA has not approved.

The approval rests on rigorous trials. In 120 children and young adults with Dravet syndrome, 20 mg/kg per day of pharmaceutical CBD for 14 weeks cut median monthly convulsive seizures from 12.4 to 5.9, versus 14.9 to 14.1 on placebo; 43% of patients at least halved their convulsive-seizure frequency versus 27% on placebo (a difference that did not reach statistical significance), at the cost of more diarrhea, vomiting, fatigue, fever, sleepiness and abnormal liver tests (Devinsky 2017). In 171 patients with Lennox-Gastaut syndrome, the same dose reduced drop seizures by a median 43.9% versus 21.8% on placebo (Thiele 2018). Note the doses: for a 70 kg adult, 20 mg/kg is 1,400 mg a day of purified CBD - a different universe from retail products.

The 2018 Farm Bill defined hemp as cannabis containing no more than 0.3% delta-9 THC by dry weight and removed it from the Controlled Substances Act, which is why hemp-derived CBD products are widely sold. The FDA, however, retains authority over those products under the Food, Drug and Cosmetic Act and holds that CBD cannot lawfully be marketed as a dietary supplement or added to food, because it is the active ingredient in an approved drug. In January 2023 the agency stated that the existing regulatory frameworks for foods and supplements are “not appropriate for cannabidiol” and that it would work with Congress on a new way forward; the FDA page, last updated in July 2024, still lists that as its position (FDA). State laws differ from federal law and from each other (Balachandran 2021). None of this is legal advice. The practical point is that the products on shelves sit in a regulatory gap, with no pre-market review of their content or dose.

The evidence for anxiety and pain

The most promising non-epilepsy evidence is for anxiety: controlled studies have found single doses around 300 to 600 mg reduced acute anxiety in laboratory settings, though everyday products usually contain far less (interactions and uses review).

The laboratory studies are consistent but small. In 24 treatment-naive people with social anxiety disorder, a single 600 mg dose taken 90 minutes before a simulated public-speaking test reduced anxiety, discomfort and cognitive impairment to the level of healthy controls (Bergamaschi 2011). Dose-ranging studies in healthy volunteers then found an inverted U: 300 mg reduced speech anxiety, while 100 mg, 150 mg, 600 mg and 900 mg did not (Zuardi 2017; Linares 2019). The often-cited “large case series” of 72 psychiatric outpatients, in which anxiety scores fell in 79% within a month, had no control group, and its sleep scores fluctuated over time (Shannon 2019).

Evidence for chronic pain and sleep is mixed and earlier-stage. The largest placebo-controlled pain trial at retail-level doses was negative: 136 patients with hand osteoarthritis or psoriatic arthritis were randomized to 20–30 mg of synthetic CBD or placebo daily for 12 weeks, and the difference in pain was 0.23 mm on a 100 mm scale - nothing - with no effect on sleep, anxiety or depression scores either (Vela 2022). That dose is typical of retail products; the doses that did something in the anxiety studies were ten to thirty times higher.

CBD is generally well tolerated, with diarrhea, fatigue, and appetite changes the most common side effects.

The interaction most people miss

CBD inhibits several liver enzymes (CYP3A4, CYP2C19, and others) that metabolize many medications - the same pathway behind the “grapefruit warning.” That means it can raise blood levels of other drugs (some blood thinners, seizure medicines, statins, and more), and high doses can elevate liver enzymes (drug-interactions review).

The clinical data behind that warning:

  • In 81 epilepsy patients started on pharmaceutical CBD, blood levels of clobazam’s active metabolite, topiramate, rufinamide, zonisamide and eslicarbazepine rose as the CBD dose rose (most stayed within the accepted therapeutic range; the clobazam metabolite did not), and liver enzymes were significantly higher in those also taking valproate (Gaston 2017).
  • In 16 healthy adults taking 1,500 mg per day for about 3.5 weeks, 7 (44%) developed raised ALT and 5 (31%) exceeded five times the upper limit of normal - meeting the consensus definition of drug-induced liver injury - with nothing at baseline predicting who (Watkins 2021).
  • The comprehensive review lists interactions with antiepileptics, antidepressants, opioid analgesics, THC, acetaminophen and alcohol (Balachandran 2021). Warfarin is a frequently raised concern because small changes in its blood level change bleeding risk.

The cautionary list, then: anticoagulants, anti-seizure drugs, sedatives and benzodiazepines, some antidepressants, immunosuppressants, statins, and anything that already carries a grapefruit warning.

Product quality: the labeling problem

  • A certificate of analysis (COA) from a third-party lab confirming CBD content and the absence of contaminants is the only document that says what is actually in a bottle. When 84 CBD products sold online were tested, only 31% were labeled accurately (within 10%); 26% contained less CBD than stated, 43% more, and THC was detected in 21% (Bonn-Miller 2017).
  • Realistic dosing - many cheap products are underdosed. Typical retail serving sizes are also far below the amounts used in the studies that reported an effect, which is one reason those results may not apply to retail products.
  • A conversation with your pharmacist if you take any prescription medication.

Nothing here is a product recommendation. The guide to reading a supplement study applies doubly to a category with this much marketing.

Who should talk to a clinician first

Anyone on prescription medication - especially blood thinners, anti-seizure drugs, sedatives or immunosuppressants; anyone with liver disease or abnormal liver tests; anyone pregnant or breastfeeding; anyone under 18; and anyone whose anxiety, pain or sleep problem is persistent enough to affect daily life, since those need assessment rather than a hemp extract. Stop and seek care for yellowing skin or eyes, dark urine or unusual fatigue.

Bottom line

CBD is genuine medicine for two rare epilepsies and a promising-but-unproven supplement for anxiety and pain. Its most underrated risk is drug interactions, not intoxication. Choose third-party-tested products and check with a clinician before combining CBD with other medications.

Related

More from SelfHacking

Supplements

Multivitamins: The COSMOS Memory Trials and What Else Holds Up

Three COSMOS sub-studies found a daily multivitamin nudged memory scores in adults over 60. How big the effect is, why an earlier trial found nothing, and where cancer, heart and lifespan data stand.

Supplements

Bovine Colostrum: Human Trials on Gut Barrier and Immunity

Human trials link bovine colostrum to lower gut permeability, fewer respiratory symptoms in hard-training athletes and modest lean-mass gains. The doses used, the limits and the safety picture.

Readers

Comments

Leave a comment

Comments are reviewed before they appear.