CBD vs THC: Real Differences According to PubMed, EMA and WHO (2026)
Share
CBD vs THC: Real Differences According to PubMed, EMA and WHO
Contents
1. The same molecule, opposite effects
CBD (cannabidiol) and THC (delta-9-tetrahydrocannabinol) are structural isomers: identical molecular formula C21H30O2, molecular weight 314.47 g/mol, different atomic arrangement. Both are 21-carbon terpenophenolic phytocannabinoids exclusive to the Cannabis genus, and both interact with the human endocannabinoid system.
Cannabidiol
Non-psychoactive
Negative allosteric modulator of CB1 (NAM)
Partial agonist 5-HT1A (serotonin)
GPR55 antagonist
TRPV1 and Ca2+ channel activity
Delta-9-tetrahydrocannabinol
Psychoactive
Partial agonist CB1 (orthosteric site)
Partial agonist CB2
Activates mesolimbic dopaminergic pathway
Positive intrinsic efficacy at CB1
The CB1 and CB2 receptors are G-protein coupled receptors (GPCRs) with 68% sequence identity in their seven transmembrane domains. CB1 is concentrated in the CNS: prefrontal cortex, hippocampus, basal ganglia, amygdala. CB2 predominates in the immune system and peripheral tissue. Source: Pertwee et al., British Journal of Pharmacology 2008, PMC2219532.
2. Why THC gets you high and CBD does not: the exact mechanism
THC is a partial agonist at the orthosteric site of CB1. Upon binding it activates dissociation of the Gi/o protein subunits:
- Gi-alpha subunit inhibits adenylyl cyclase: decrease in intracellular cAMP
- Gi-betagamma subunits open GIRK channels: membrane hyperpolarisation
- CB1 activation in mesolimbic dopaminergic pathway: dopamine release in nucleus accumbens, euphoria and reinforcing effect
- CB1 activation in hippocampus: inhibition of LTP (long-term potentiation), short-term memory impairment
- CB1 activation in amygdala: anxiety, paranoia or euphoria depending on dose and context
CBD binds to an allosteric site different from the orthosteric site of CB1 (it does not compete with THC). Negative allosteric modulators by definition have no intrinsic efficacy of their own: in the absence of agonists, CBD has minimal effect on basal endocannabinoid tone. Without orthosteric CB1 activation, there is no mesolimbic dopamine release and no psychoactivity.
Site-directed mutagenesis analysis and molecular dynamics confirm that CBD binds to the N-terminal domain of CB1 and allosterically couples towards the orthosteric site, reducing the receptor responsiveness. Sources: PMC6650144; PMC8122825.
Cannabidiol should not be subject to international scheduling. It has no psychoactive properties and presents no potential for abuse or dependence.
Source: WHO ECDD CBD Critical Review Report, November 2018 (cdn.who.int).
Additionally, CBD acts on other pharmacological targets: 5-HT1A receptor (partial agonist, anxiolytic implications), TRPV1 (agonist, desensitisation), GPR55 (antagonist, cell proliferation), voltage-dependent calcium channels (inhibition, anticonvulsant activity). Source: PMC10458707.
3. Comparative pharmacokinetics
| Parameter | CBD | THC |
|---|---|---|
| Oral bioavailability | ~6% | 4–12% |
| Inhaled bioavailability | ~31% | 10–27% |
| Half-life inhaled | 27–35 hours | ~4.3 days (chronic users) |
| Main metabolism | CYP2C19, CYP3A4 to 7-OH-CBD (inactive) | CYP2C9, CYP3A4 to 11-OH-THC (active) to THCCOOH (inactive) |
| Psychoactive metabolite | No | Yes: 11-OH-THC (more potent in some effects) |
| Urine detection | Days | Weeks (THCCOOH in chronic users) |
| Excretion | Faeces and urine (metabolites) | ~65% faeces, ~20% urine in 5 days |
Both molecules are highly lipophilic and widely distributed in adipose tissue. Prolonged THC detection in urine via THCCOOH does not correlate with current functional impairment. Sources: PMC6275223; PMC2689518.
THCCOOH (inactive THC metabolite) is extremely lipophilic and accumulates in adipose tissue, releasing slowly over weeks. This explains prolonged detection in chronic users, with no correlation to current functional impairment. Most urine drug tests do not detect CBD or its metabolites.
4. EMA-approved medicines based on CBD and THC
1. Lennox-Gastaut Syndrome (LGS): adjunctive therapy with clobazam. Efficacy: up to 64% reduction in atonic seizures vs 31% placebo.
2. Dravet Syndrome (DS): adjunctive therapy with clobazam. Efficacy: up to 61% reduction in convulsions vs 38% placebo.
3. Tuberous Sclerosis Complex (TSC, 2021 variation): 49% reduction vs 27% placebo.
Dosage: start 2.5 mg/kg/12h, increase to 5 mg/kg/12h, maximum 10–12.5 mg/kg/12h by indication.
EMA note: the exact mode of action is not completely understood. Calcium channels and adenosine signalling have been proposed.
5. Documented adverse effects
— Meta-analysis Schizophrenia Bulletin 2016 (PMID 26884547): OR 3.90 (95% CI 2.84–5.34) for more frequent users vs non-users of developing psychosis. OR 1.97 for moderate users.
— EU-GEI study The Lancet Psychiatry 2019 (S2215-0366(19)30048-3): daily use of high-potency cannabis (>10% THC) associated with OR approximately 5x greater for psychosis. Prospective, multicentre design, 11 European countries.
THC dependence (NIDA data, National Institute on Drug Abuse): approximately 9% of all cannabis users develop Cannabis Use Disorder (CUD). Risk increases to 17% if use begins in adolescence, and to 20–30% in daily users. Documented withdrawal syndrome: irritability, insomnia, anxiety (onset 1–3 days, peak 2–6 days).
In Epidyolex clinical trials: transaminase (ALT/AST) elevations above 3× the upper limit of normal, especially when CBD is administered concomitantly with valproate. The EMA requires hepatic monitoring. The FDA issued a Drug Safety Communication in 2019.
CBD inhibits CYP2C9, CYP2C19, CYP2D6 and CYP3A. Clinical relevance is significant mainly at doses above 300 mg/day. Sources: PMC6678684; PMC11079547.
Most frequent adverse effects according to EMA summary of product characteristics (Epidyolex, more than 1 in 10 patients): somnolence, decreased appetite, diarrhoea, fever, fatigue, vomiting.
6. European legal framework 2026
CBD as Novel Food (EU Regulation 2015/2283): since January 2019, the European Commission classifies Cannabis sativa extracts containing CBD as novel food. Basis: no Member State was able to demonstrate a history of significant consumption before 15 May 1997 (the regulation cut-off date). This means products with CBD require prior authorisation before being marketed as food or supplements in the EU.
The EFSA NDA Panel published on 9 February 2026 a provisional safe level of 0.0275 mg per kilogram of body weight per day, equivalent to approximately 2 mg/day for a 70 kg adult. Applies only to food supplements with CBD of purity ≥98%, without nanoparticles.
EFSA also concludes that CBD safety cannot be established for under-25s, pregnant and breastfeeding women, and people on medication. Source: EFSA Journal 2026, DOI 10.2903/j.efsa.2026.9862.
CJEU ruling C-663/18 (Kanavape case, November 2020): the Court of Justice of the EU ruled that CBD extracted from the whole cannabis plant is NOT a narcotic under the 1961 Single Convention on Narcotic Drugs. Whole-plant CBD is subject to the principle of free movement of goods in the EU, provided it complies with food law.
THC limit in industrial hemp: the EU sets a maximum of 0.2% THC in industrial hemp varieties for the CAP (Common Agricultural Policy). The debate on raising this limit to 0.3% continues in 2026.
7. Full comparison table
| Feature | CBD | THC |
|---|---|---|
| Molecular formula | C21H30O2 (identical — structural isomers) | |
| Action on CB1 | NAM (allosteric site) | Partial agonist (orthosteric site) |
| Psychoactivity | No | Yes |
| Euphoria | No | Yes |
| Abuse potential | No (WHO 2018) | Yes: CUD in 9% of users |
| Psychosis risk | Not documented | OR 3.9–5x in frequent users (Lancet, SB) |
| Hepatotoxicity | Yes at high doses or with valproate | Low at therapeutic doses |
| CYP450 inhibition | Yes with >300 mg/day | Minor |
| EMA medicine | Epidyolex (3 indications) | Sativex with CBD (~28 countries) |
| EU food legal status | Novel Food (active process) | Narcotic under 1961 Convention |
| Urine detection | Days | Weeks (THCCOOH) |
Verified sources
- PMC2219532 — Pertwee et al., CB1/CB2 pharmacology, British Journal of Pharmacology 2008
- PMC6275223 / PMID 30534073 — Systematic review CBD pharmacokinetics in humans
- PMC6650144 — CBD as NAM of CB1 in silico with THC
- PMC8122825 — Allosteric modulation of CB1 by CBD, molecular modelling
- PMC10458707 — Molecular and cellular mechanisms of CBD
- PMID 26884547 — Meta-analysis OR cannabis psychosis, Schizophrenia Bulletin 2016
- S2215-0366(19)30048-3 — EU-GEI study, The Lancet Psychiatry 2019
- PMC6678684 — CBD and CYP450 interactions, adverse effects
- PMC11079547 — Drug-drug interactions medical cannabis
- EMEA/H/C/004675 — Epidyolex full EPAR (ema.europa.eu)
- WHO ECDD 2018 — CBD Critical Review Report (cdn.who.int)
- EFSA Journal 2026, DOI 10.2903/j.efsa.2026.9862 — Provisional safe level CBD
- CJEU C-663/18 — Kanavape case, November 2020
- PMC2689518 — Human Cannabinoid Pharmacokinetics
- NIDA — Cannabis and Marijuana (nida.nih.gov)
Beetle Print — Custom accessories for the sector
Grinders, rolling trays, Clipper lighters and custom packs for headshops, grow shops and associations across Europe.
VIEW COLLECTION