Between completing my senior thesis and my planned move to Seattle, I suddenly panicked and wondered: Is this truly the career path for me? Ultimately, I realized I did not have what it took to be a basic science researcher, so I worked for a year as a biochemical technician and applied to medical school—a decision I have never regretted.
Yet, I do not regret my background in biochemistry either. It is the foundation upon which all I have learned of medicine and psychiatry is built. The field of neuropsychopharmacology has evolved over the past 30 years, and I am filled with pride, awe, and excitement to have witnessed the clinical advancements that basic science researchers have made possible. However, in my opinion, there is 1 significant exception to all my accolades for how good science has advanced medicine: the research on cannabis.
The scientific method is used in medicine and has been the established paradigm for scientific inquiry since the 1600s. Thus, all clinical trials submitted to the US Food and Drug Administration (FDA) are meticulously designed to answer a primary outcome question about the drug under investigation. A necessary requirement of the study design is to ensure that the variables in the various arms of the study are identical except for one: drug A at dose X compared with placebo. If multiple doses of drug A or multiple drugs are being compared, each arm is precisely characterized so the conclusions drawn from the primary outcome will be highly accurate and reliable.
Components of Cannabis
Over the past 10 years, I have read numerous articles in the psychiatric literature that draw conflicting conclusions about the effect of cannabis on cognition, its attendant risk of psychosis, its addictive potential, and its effect on suicidality, and about its impacts on depression, anxiety, pain relief, and apathy. (There appears to be only 1 consistent finding: Heavy cannabis use during brain development increases the risk of psychosis and cognitive impairment, especially when other risk factors are present.) A likely explanation for the wide-ranging discrepancies is that cannabis is the product of a plant that is not a single molecule or pure substance. Many articles with the word cannabis in the title can be discarded as meaningless unless the authors took the additional step of testing all of the cannabis used by their study subjects for the quantitative and qualitative molecules present.
It is well established that cannabis contains more than 500 unique molecular constituents.2,3 These constituents include more than 100 cannabinoids, more than 200 terpenoids (terpenes), and numerous flavonoids. Two cannabinoids are most prevalent in cannabis: Δ-9-tetrahydrocannabinol (THC) and cannabidiol (CBD), both of which exist in an extremely wide range of ratios depending on the specific strain of cannabis assayed. Significantly, these 2 cannabinoids are oppositional in their clinical effects and have dramatically different pharmacodynamic effects. In reviewing the literature on cannabis, a consistent finding is that potency of cannabis is a proxy for the amount of THC. Although this is likely a result of the psychoactive effects of cannabis that have been noted for more than 6000 years,4 CBD has surfaced as an important cannabinoid with many positive medicinal properties. Notably, cannabis remains a Schedule 1 drug by the Drug Enforcement Administration (DEA), while CBD is unscheduled.
The Table documents some of the significant differences between THC and CBD. Despite these differences, the vast majority of publications on cannabis draw conclusions based on participants’ self-reports of quantity and frequency of cannabis use. In these studies, the molecular components of the cannabis used are not known, nor is the important ratio of THC to CBD.
A fair analogy would be to evaluate a new drug; let us call it Sedalant. Sedalant contains a wide-ranging ratio of 2 molecules with different mechanisms of action: one a sedative, and one a stimulant. Without knowing the exact ratios and concentrations of each of these molecules in clinical studies of Sedalant, no evidence-based conclusions could be drawn from the drug studies.
Naturally occurring cannabis plants produce roughly an equal ratio of THC to CBD, containing approximately 2% of each. However, over the decades, with selective breeding in some strains, THC has been increased significantly while CBD has been virtually bred out. ElSohly et al5 analyzed the concentrations of THC and CBD from 38,681 samples of illegal cannabis seized by the DEA between 1995 and 2014. The authors found a clear trend of increased THC in cannabis over time, with an average concentration of 4% in 1995 that increased to 12% in 2014. Simultaneously, CBD concentration steadily declined, from 0.5% in 2004 to less than 0.2% in 2014. The investigators concluded that the ratio of THC to CBD increased from 14-fold in 1995 to 80-fold in 2014.5 Some strains of cannabis today may contain up to 30% THC with virtually no CBD, while other strains have high concentrations of CBD with minimal THC.
To further complicate the cannabis story, very few of the other cannabinoids in cannabis plants have been characterized pharmacodynamically or pharmacokinetically. The characterization of 3 of these more than 100 cannabinoids—Δ-9-tetrahydrocannabivarin, cannabichromene, and cannabigerol—tells us that they are pharmacologically distinct from both THC and CBD.4,6,7
Interestingly, the odors (skunk is a common description) often associated with cannabis are largely due to the complex combination of terpenes, another compound that complicates the study of cannabis. (Cannabinoids are odorless.) An extensive literature documents the essential role terpenes play in the biology of the cannabis plant, the plant’s production of cannabinoids, and the properties of a specific strain of cannabis that help determine its market value. Terpenes likely also contribute pharmacodynamic and pharmacokinetic effects that impact the neuropsychopharmacology resulting from ingestion of each different strain of cannabis.8,9
Several publications since 2018 have appropriately concluded that we know very little about the range of pharmacological effects from cannabis due to the large variability of cannabinoids, terpenes, and flavonoids that are present in different strains.9-11
For this very reason, it would be contrary to the scientific method of research to attempt to gain FDA approval of cannabis.
Only 1 of the more than 100 cannabinoid components in cannabis has been sufficiently studied to receive FDA approval: CBD. Following a comprehensive analysis as well as clinical trials of pure CBD, it received approval in 2018 for the treatment of seizures in Dravet syndrome and in Lennox-Gastaut syndrome in children and adults.12 We may see additional approvals in the future for this well-studied compound, as a growing and consistent body of evidence shows that pure CBD in defined dosages has antipsychotic effects in patients with schizophrenia.13
The practice of medicine is built upon evidence-based treatments that have been appropriately vetted by well-established protocols. CBD traveled this path to become an FDA approved drug, and so should any other component of the 500-plus molecules in cannabis. Accordingly, clinical publications should be held to a higher standard of identifying and defining what components of cannabis are relevant in any article being submitted for publication.
With each passing year, more US states legalize cannabis for medical use, recreational use, or both. It is likely that federal laws will eventually make cannabis use legal on a national level. That will create an opportunity to require a comprehensive analysis and disclosure of all the components of the numerous strains of cannabis currently being sold in state dispensaries or on the streets. Concurrently, the individual molecular components of the cannabis plant should be identified and pharmacologically characterized. This would create a treasure trove of possible drug candidates for specific clinical indications, and it would allow for a more informed choice when choosing a cannabis product.
In this scenario, cannabis should be regulated federally similar to tobacco products and alcohol. Both of these have significant risks, and both are tightly controlled. Ideally, the taxes imposed on federally regulated cannabis could help fund the extensive research needed to allow a more competent understanding of the complex combination of molecules that different strains of cannabis contain.
1. The word cannabis should not be used casually in medical literature, as it is nonspecific and includes varying molecular combinations of diverse cannabinoids, terpenes, and flavonoids.
2. Future research involving cannabis should use a single strain in each study; each strain used should be analyzed quantitatively and qualitatively for biologically active components.
3. If multiple strains are used in a clinical trial, they should be analyzed (as noted in #2), and each strain should be studied as a separate arm compared with placebo.
4. Increased funding is needed to accurately characterize the 500-plus molecular components of the cannabis plant, including the pharmacokinetics and pharmacodynamics of each component.
5. A public education campaign should aggressively disseminate the molecular facts about cannabis, especially the oppositional effects of THC and CBD. Similarly, the risks of increased psychosis and cognitive impairment with heavy regular use of THC in the developing brain should be explained, much like disclaimers on alcohol about drinking during pregnancy.
6. Cannabis should be federally legalized and regulated similarly to alcohol and tobacco products.
7. Cannabis should not be regulated by the FDA due to its extreme heterogeneity of components and the associated unpredictable pharmacological properties.
8. Cannabis should not be prescribed by medical practitioners. However, as molecular components of cannabis are developed and garnish FDA approval, those components should be prescribed appropriately.
9. Pharmacological exploration and development of cannabis components that demonstrate medical benefits should be continued. CBD is a successful model of such.
10. Medical practitioners should discuss cannabis with their patients only after they have attained competence in understanding the science and factual risks/benefits/adverse effects of cannabis, similar to how a practitioner would discuss alcohol or tobacco use.
With this structure in place, we have the ability to inform and support our patients appropriately. And, with more information, we can help end the cannabis confusion.
Dr Miller is medical director, Brain Health, Exeter, New Hampshire; Editor in Chief, Psychiatric TimesTM; staff psychiatrist, Seacoast Mental Health Center, Exeter; Consulting Psychiatrist, Exeter Hospital, Exeter; Consulting Psychiatrist, Insight Meditation Society, Barre, Massachusetts.
1. Miller J. An Analysis of the Fate of the Escherichia coli 4.5S RNA Gene in Yeast Using a Recombinant Plasmid. Senior undergraduate honors thesis. University of Washington, Amherst; 1981.
2. Lafaye G, Karila L, Blecha L, Benyamina A. Cannabis, cannabinoids, and health. Dialogues Clin Neurosci. 2017;19(3):309-316.
3. Pertwee R, ed. Handbook of Cannabis. Oxford University Press; 2014.
4. Atakan Z. Cannabis, a complex plant: different compounds and different effects on individuals. Ther Adv Psychopharmacol. 2012;2(6):241-254.
5. ElSohly MA, Mehmedic Z, Foster S, et al. Changes in cannabis potency over the last 2 decades (1995-2014): analysis of current data in the United States. Biol Psychiatry. 2016;79(7):613-619.
6. ElSohly MA, Radwan MM, Gul W, et al. Phytochemistry of Cannabis sativa L. Prog Chem Org Nat Prod. 2017;103:1-36.
7. Nachnani R, Raup-Konsavage WM, Vrana KE. The pharmacological case for cannabigerol. J Pharmacol Exp Ther. 2021;376(2):204-212.
8. Sommano SR, Chittasupho C, Ruksiriwanich W, Jantrawut P. The cannabis terpenes. Molecules. 2020;25(24):5792.
9. Booth JK, Bohlmann J. Terpenes in cannabis sativa – from plant genome to humans. Plant Sci. 2019;284:67-72.
10. Baron EP. Medicinal properties of cannabinoids, terpenes, and glavonoids in cannabis, and benefits in migraine, headache, and pain: an update on current evidence and cannabis science. Headache. 2018;58(7):1139-1186.
11. Jett J, Stone E, Warren G, Cummings KM. Cannabis use, lung cancer, and related issues. J Thorac Oncol. 2018;13(4):480-487.
12. Epidiolex. Greenwich Biosciences; 2020. Updated April 2020. Accessed July 7, 2021.
13. McGuire P, Robson P, Cubala WJ, et al. Cannabidiol (CBD) as an adjunctive therapy in schizophrenia: a multicenter randomized controlled trial. Am J Psychiatry. 2018;175(3):225-231.
14. Elsaid S, Le Foll B. The complexity of pharmacology of cannabidiol (CBD) and its implications in the treatment of brain disorders. Neuropsychopharmacol. 2020;45(1):229-230.
15. National Academies on Sciences, Engineering, and Medicine. The Health Effects of Cannabis and Cannabinoids: The Current State of Evidence and Recommendations for Research. The National Academies Press; 2017.
16. Rong C, Carmona NE, Lee YL, et al. Drug-drug interactions as a result of co-administering Δ9-THC and CBD with other psychotropic agents. Expert Opin Drug Saf. 2018;17(1):51-54. ❒