This essay seeks to provide answers to the following questions:
1. To what extent do dose, age, and genetic background transform cannabis from a potentially useful therapeutic agent into a driver of psychosis and other long‑term harms?
2. How should clinical practice and public policy adapt to high‑potency, widely marketed cannabinoids (THC, CBD, and beyond) without either demonizing all use or trivializing genuine risks?
Cannabis is increasingly portrayed as a relatively benign or even “natural” remedy, especially when compared with alcohol or opioids, while commercial forces promote high‑potency products and recent regulatory changes implicitly signal that frequent use is innocuous. This narrative stands in tension with a substantial body of longitudinal, genetic, and toxicological evidence. Large cohort studies demonstrate that cannabis use, particularly when initiated in adolescence, sustained over time, and involving potent THC formulations, is associated with elevated risks of psychotic outcomes, earlier onset of schizophrenia‑spectrum disorders, and measurable cognitive and functional impairment. Genetic susceptibility, exemplified by the AKT1 rs2494732 polymorphism, further amplifies risk in sizeable subgroups of the population. In parallel, experimental work reveals that ostensibly benign cannabinoids such as CBD and CBDV can induce DNA damage and chromosomal aberrations in human‑derived cells at concentrations consistent with consumer‑attained levels, raising concerns about long‑term genotoxicity. While cannabinoids do have legitimate therapeutic roles when carefully dosed and medically supervised, prevailing market practices, such as high doses, rapid‑onset formulations, youth‑oriented branding especially of non-regulated technocannabinoids, diverge markedly and uncontrolably from evidence‑based regimens. This essay argues for a recalibration of the risk / benefit discussion, emphasizing delayed initiation beyond vulnerable developmental windows, cautious dosing and micro‑dosing strategies where appropriate, attention to genetic and psychiatric risk factors, and the need for robust regulatory and educational frameworks. Prescription privileges, it is suggested, should be limited to clinicians with dedicated training and continuing medical education in cannabinoid pharmacology; public‑health policy should be guided by independent evidence rather than commercial narratives.
Cannabis use is associated with significant, dose‑dependent health risks, particularly for adolescents and genetically susceptible individuals, and these risks extend beyond THC to include potential genotoxic effects of widely used “non‑psychoactive” cannabinoids such as CBD.
Public policy and commercial messaging increasingly frame cannabis as relatively benign, especially when compared with alcohol or opioids, and recent moves to reschedule cannabis in most jurisdictions reinforce the perception of reduced danger. At the same time, potency has risen sharply, and high‑THC products, concentrates, and edibles routinely deliver doses far beyond those studied in older epidemiologic cohorts. While selected medical uses, such as low‑dose THC for chemotherapy‑induced nausea, are well supported, the broader push to normalize frequent, high‑dose, and adolescent use overlooks a substantial body of evidence linking cannabis exposure to psychosis, cognitive impairment, and, in the case of CBD and related cannabinoids, genotoxicity in human‑derived cells (Andréasson S. et al. 1987; Zammit S. et al. 2002; Russo C. et al. 2019).
Large longitudinal cohort studies provide some of the strongest evidence connecting cannabis use with later psychotic outcomes. The classic Swedish conscript study followed approximately 45,570 young men for 15 years and found that heavy cannabis users (more than 50 lifetime uses) had a six‑fold increased risk of developing schizophrenia compared with non‑users, even after adjustment for social background and other psychiatric illness. Subsequent re‑analyses confirmed that the association persisted when restricting to schizophrenia cases emerging years after baseline, arguing against simple reverse causation (i.e., prodromal individuals self‑medicating with cannabis) (Andréasson S. et al. 1987; Zammit S. et al. 2002).
Prospective birth‑cohort studies from New Zealand, including the Dunedin and Christchurch Health and Development Study (CHDS) cohorts, extended these findings to community samples with repeated assessments. These cohorts showed that adolescent and early‑adult cannabis use predicted increased rates of psychotic symptoms and schizophrenia‑spectrum diagnoses in adulthood, even after controlling for childhood psychotic‑like experiences and multiple social confounders. The effect was stronger with earlier onset and with persistent use, supporting a causal contribution of cannabis exposure to psychosis risk rather than mere correlation (Arseneault L. et al. 2002).
Both observational cohorts and meta‑analytic syntheses indicate that psychosis risk rises in a dose‑ and frequency‑dependent manner: more frequent use, higher cumulative exposure, and more potent THC formulations are associated with greater relative risk. Meta‑analyses encompassing dozens of studies report that any cannabis use is linked to an approximately 30% increase in psychosis risk, with substantially higher odds among daily or high‑potency users. Importantly, a large meta‑analysis of over 80 studies found that the age of onset of psychosis in cannabis users occurred on average 2.7 years earlier than in non‑users, underscoring the concern that cannabis can precipitate or accelerate psychotic disorders in vulnerable individuals.
Neurodevelopmental timing appears critical because the human brain continues to mature into the mid‑20s, with ongoing synaptic pruning, myelination, and refinement of fronto‑striatal circuits implicated in cognition and emotional regulation. Exogenous cannabinoids acting on the endocannabinoid system during adolescence may perturb these processes, altering dopaminergic signaling and cortical connectivity in ways that increase vulnerability to persistent psychotic illness. This developmental context makes adolescence and young adulthood the highest‑risk window and supports strong public‑health messaging to delay initiation and avoid heavy use during these years (find extensive bibliography here: Gardner, A. 2023).
Genetic susceptibility and AKT1 gene
Inter‑individual differences in psychosis risk are partly explained by common genetic variants that modulate the response of the brain to THC. Among these, the AKT1 rs2494732 polymorphism has received particular attention: carriers of two copies of the “C” allele have been reported to exhibit approximately a seven‑fold increase in psychosis risk among regular cannabis users compared with non‑carriers, suggesting a gene–environment interaction. AKT1 (PKBα) participates in intracellular signaling pathways linked to dopamine neurotransmission, and the risk allele may prolong or amplify THC‑induced dopamine elevation, a state closely tied to the emergence of psychotic symptoms.
Beyond its dopaminergic roles, AKT1 (PKBα) is also a central survival kinase in the phosphatidylinositol‑3‑kinase (PI3K)/AKT pathway, exerting anti‑apoptotic effects in many cell types. Cannabinoids can bidirectionally influence this axis: at low concentrations, CB1/2 receptor activation can rescue cells from apoptosis via PI3K–AKT and ERK‑mediated pro‑survival signaling, whereas at higher, cytotoxic concentrations they can induce AKT inhibition through de novo ceramide synthesis in cytoplasm, promoting apoptosis in tumor cells, probably in a CB1/2-independent way (Galve-Roperh 2002; Gómez Del Pulgar 2002). Although AKT1 rs2494732 has so far been studied in relation to psychosis rather than cancer, the convergence of AKT1‑dependent survival signaling and dose‑dependent, CB‑receptor‑linked versus ceramide‑mediated pro‑apoptotic effects of cannabinoids suggests a plausible, though as yet untested, interface between AKT1 genotype, low‑dose cannabinoid exposure, and cancer cell survival or promotion. This remains a hypothesis‑generating consideration rather than an established clinical risk, but it warrants caution and further investigation on the part of cannabinoid prescribers.
Epidemiologic data indicate that the minor allele frequency (MAF) at rs2494732 is high (on the order of 30–35%), meaning that a large fraction of the population carries at least one copy of the risk allele, even though only homozygous carriers appear to be at the highest risk. In practice, most users have no information about their AKT1 status, yet they are exposed to increasingly potent THC products marketed aggressively, including to young adults. This combination (common high‑risk genotypes, high‑THC formulations, and early‑onset use) creates a substantial, largely invisible risk pool that standard public messaging often fails to address (see Footnote 1).
While THC rightly receives much of the attention in discussions of neuropsychiatric risk, the rapid expansion of cannabidiol (CBD) products has created an impression that non‑psychotropic cannabinoids (see Footnote 2) are uniformly safe. However, in vitro and in vivo toxicological work challenges this assumption. The study by Chiara Russo et al. (2019) investigated the genotoxic potential of cannabidiol (CBD) and its propyl analogue cannabidivarin (CBDV) in human‑derived cells under exposure conditions intended to approximate those experienced by consumers.
Using single‑cell gel electrophoresis in human liver (HepG2) and buccal‑derived (TR146) cell lines, the authors found that both CBD and CBDV induced DNA damage at low concentrations, beginning at approximately 0.2 µM. The same research group and related work also reported chromosomal aberrations and micronucleus formation in animal models exposed to these cannabinoids, and they argue that fixation of DNA damage in the form of chromosomal alterations is a recognized step in multistage carcinogenesis, implying potential carcinogenic properties with chronic exposure. While in vitro genotoxicity does not automatically translate into clinical cancer risk, these findings directly contradict the widely held assumption that “hemp‑derived” cannabinoids are harmless at consumer‑relevant doses and underscore the need for cautious, long‑term safety evaluation of high‑dose, chronic CBD use.
The issue of low vs. high cannabinoid dosing in malignant disease has been previously addressed (see Cannabis and Cancer). Given that the incidence of occult neoplasia in random autopsies is on the order of 4–8% (Uozaki H. et al. 2026), the risk of exacerbating an occult tumor by administering low doses of cannabinoids for a not seriously considered indication warrants careful consideration (primum non nocere).
Beyond psychosis and genotoxicity, cannabis use—particularly when initiated in adolescence and sustained at moderate to high doses—is associated with a range of cognitive and functional consequences. Longitudinal cohort data from New Zealand and other settings have linked persistent cannabis use beginning in adolescence with measurable decrements in neuropsychological performance, including memory, attention, and executive function, some of which appear only partially reversible with abstinence. Functional outcomes such as lower educational attainment, reduced occupational performance, and increased risk of depressive symptoms have also been reported, though disentangling causality from shared social and genetic factors remains challenging (Arseneault L. et al. 2002).
From a systemic perspective, inhaled combustion products from smoked cannabis share many of the respiratory risks seen with tobacco, including chronic bronchitis symptoms and airway inflammation; although the relationship to lung cancer remains less clear, the presence of polycyclic aromatic hydrocarbons and other carcinogens supports a precautionary stance. Cardiovascular effects, such as transient tachycardia, blood pressure changes, and a possible increase in short‑term risk of myocardial infarction in predisposed individuals, have been described, particularly with high‑dose THC. Taken together, these data situate cannabis not as a uniquely dangerous drug in absolute terms, but as a substance with a non‑trivial, multidimensional risk profile that escalates with dose, chronicity, mode of administration, and host susceptibility (indicative article: Nelson K.M. et al. 2020).
Despite these risks, cannabinoids have genuine therapeutic applications when used judiciously and at carefully titrated doses. THC‑containing preparations are effective for chemotherapy‑induced nausea and vomiting and for certain forms of refractory appetite loss, and CBD has proven utility in specific epileptic encephalopathies when administered under medical supervision. The key concept is that “the dose makes the poison”: low, intermittent doses tailored to a defined indication can deliver benefit with acceptable risk, whereas high, frequent, recreational doses—especially in young or vulnerable individuals—shift the balance decisively toward harm (see also Cannabis and Cancer and Nelson K.M. et al. 2020).
Clinical and translational literature increasingly supports micro‑dosing strategies for some indications, particularly benign and neurodegenerative conditions (Ruver‑Martins A.C. et al. 2022), where the minimum effective dose is identified and upward titration is avoided, rather than the high‑dose, rapid‑onset culture promoted by many commercial products. For both plant and purified cannabinoids, route of administration, comorbidities, concomitant medications, and genetic background (e.g., AKT1 variants, polymorphisms affecting metabolism) should inform prescribing and self‑use; yet current retail environments rarely incorporate such individualized risk assessment.
Regulatory changes that relax scheduling or expand access without aligning labeling and marketing rules to actual risk may inadvertently signal that cannabis is “safe,” particularly to adolescents and parents. Simultaneously, industry incentives favor high‑potency, maximal‑dose products, including edibles containing 50–100 mg THC per unit—doses far exceeding those associated with usual therapeutic efficacy for benign conditions and far removed from the lower potencies prevalent in earlier epidemiologic cohorts. This divergence between evidence‑based dosing and market offerings represents a public‑health concern rather than a mere consumer‑choice issue.
Effective policy should integrate several elements: clear, age‑stratified risk communication; robust restrictions on youth‑targeted marketing; and support for independent, non‑industry‑funded research on both THC and non‑THC cannabinoids, including CBD and CBDV. Gradual incorporation of pharmacogenetic testing into clinical and preventive frameworks, as such testing becomes more widely available, could enable more individualized counseling, although ethical, privacy, and access issues must be carefully addressed. Potency‑based taxation has been proposed by some (Xu L. et al. 2025) as a means to discourage ultra‑high‑THC products; the present author is of the opinion that such measures may ultimately foster increased adolescent criminality rather than mitigate the use of these products and of technocannabinoids.
Taken together, the evidence base portrays cannabis as a pharmacologically active plant whose major constituents influence neurodevelopment, dopaminergic signaling, and genomic integrity in ways that can produce adverse outcomes, sometimes serious, in various subsets of users. Psychosis risk increases with dose, frequency, potency, and early onset; genetic factors such as AKT1 variants amplify this risk; and even ostensibly benign cannabinoids like CBD and CBDV demonstrate genotoxic effects in human‑derived cells under exposure conditions relevant to consumers.
A scientifically grounded approach to cannabis must therefore move beyond both extremes of discourse—neither demonizing all use nor trivializing genuine harms—and instead emphasize cautious dosing; delayed initiation beyond vulnerable developmental windows; explicit recognition of genetic, psychiatric, and developmental risk factors; and rigorous long‑term safety evaluation of the full spectrum of cannabinoids now in widespread use. Prescription privileges should be reserved for physicians with appropriate training and relevant, up‑to‑date continuing medical education in cannabinoid pharmacology and clinical practice.
The frequency of two copies of the AKT1 rs2494732 C allele (CC homozygous genotype) in the general Caucasian population is approximately 90,000 to 120,000 per 1,000,000 individuals, or roughly 10%. This estimate derives from a minor allele frequency (MAF) of the C allele at ~0.30-0.35 in European ancestries, using Hardy-Weinberg equilibrium: frequency of CC ≈ (MAF)^2, or about 0.09-0.12. Studies in Caucasian cohorts (e.g., UK/European samples) implicitly align with this range, though exact genotype counts are rarely tabulated outside genomic databases like gnomAD. These estimates should be viewed as provisional, not exact.
The present author uses the terms psychoactive and psychotropic differently: in this respect, CBD is considered psychoactive because of its modulating effect on cognition and mood [via 5-HT1A agonism, AEA gate (Papastavrou 2024, p. 59] but not psychotropic because it does not produce any disruptive effects such as intoxication.
Andréasson, S. et al. (1987). Cannabis and schizophrenia. A longitudinal study of Swedish conscripts. Lancet. 1987 Dec 26;2(8574):1483-6. doi: 10.1016/s0140-6736(87)92620-1.
Arseneault, L. et al. (2002) Cannabis use in adolescence and risk for adult psychosis: longitudinal prospective study. BMJ. 2002 Nov 23;325(7374):1212-3. doi: 10.1136/bmj.325.7374.1212.
Galve-Roperh, I. et al. (2002). Mechanism of Extracellular Signal-Regulated Kinase Activation by the CB1 Cannabinoid Receptor. Mol Pharmacol 62:1385-1392. doi: 10.1124/mol.62.6.1385
Gómez Del Pulgar, T. et al. (2002). Cannabinoids protect astrocytes from ceramide-induced apoptosis through the phosphatidylinositol 3-kinase/protein kinase B pathway. J Biol Chem. 2002 Sep 27;277(39):36527-33. doi: 10.1074/jbc.M205797200
Large, M. et al. (2011). Cannabis use and earlier onset of psychosis: a systematic meta-analysis. Arch Gen Psychiatry. 2011 Jun;68(6):555-61. doi: 10.1001/archgenpsychiatry.2011.5
Nelson, KM., et al. (2020). The Essential Medicinal Chemistry of Cannabidiol (CBD). J Med Chem, 63:21; https://pubs.acs.org/doi/10.1021/acs.jmedchem.0c00724.
Papastavrou, A-T. (2024). Endocannabinoid system & Cannabis. Lefko Melani (Eds). ISBN: 978-618-5801-24-3. (Παπασταύρου Α-Θ. (2024). Ενδοκανναβινοειδές σύστημα & Κάνναβη. Λευκό Μελάνι (Εκδ.). ISBN: 978-618-5801-24-3.
Russo, C. et al. (2019). Low doses of widely consumed cannabinoids (cannabidiol and cannabidivarin) cause DNA damage and chromosomal aberrations in human-derived cells. Arch Toxicol. 2019 Jan;93(1):179-188. doi: 10.1007/s00204-018-2322-9.
Ruver-Martins, AC., et al. (2022). Cannabinoid extract in microdoses ameliorates mnemonic and nonmnemonic Alzheimer's disease symptoms: a case report. J Med Case Rep. 2022 Jul 12;16(1):277. doi: 10.1186/s13256-022-03457-w.
Uozaki, H. et al. (2026). Trends in the Hidden Burden of Cancer in an Autopsy-Based Study Over 66 Years in Japan. JAMA Netw Open. 2026 Feb 2;9(2):e2557812. doi: 10.1001/jamanetworkopen.2025.57812.
Xu, L., et al. (2025) How tax structures for retail cannabis shape cannabis use among youth and young adults: evidence from a volumetric choice experiment. Eur J Health Econ. 2025 Dec 29:10.1007/s10198-025-01875-3. doi: 10.1007/s10198-025-01875-3