Inventing an Addiction: The Flawed Science of Cannabis Use Disorder
The Myth of the Monolith: Why It Is Time to Dismantle "Cannabis Use Disorder"
The central premise of psychiatric nosology is utilitarian: we classify pathologies to understand their mechanisms and engineer targeted, effective interventions.
A diagnostic category earns its clinical validity only when it delineates a distinct physiological or psychological architecture that responds predictably to standardized protocols. When a construct fails this fundamental test—grouping wildly divergent biological responses under an indiscriminate behavioral label while producing interventions that perform no better than the passage of time—it ceases to be medicine. It becomes administrative theater.
"Cannabis Use Disorder" (CUD) as codified in modern diagnostic manuals has reached precisely this epistemic failure point.
By reducing an intricately polygenic, individualized neurochemical dialogue to a blunt checklist of behavioral friction, the diagnosis obscures biology, pathologizes constitutional variation, and sustains an ineffective clinical architecture. To advance precision mental health, we must dismantle Cannabis Use Disorder as a standardized clinical entity.
The Genomic Chasm: The Fallacy of Uniform Pharmacology
The diagnostic criteria for CUD presume a predictable, uniform pharmacological agent meeting a standardized neurochemical receptor system. This assumption collapses under modern pharmacogenomics and molecular pharmacology.
The human endocannabinoid system (ECS) does not operate like the relatively uniform, rigid pathways targeted by classical xenobiotics. Baseline endocannabinoid tone, retro-signaling density, and receptor affinity vary profoundly across individuals based on specific genomic signatures:
Receptor Architecture: Single nucleotide polymorphisms (SNPs) in the CNR1 gene alter the expression and conformation of the presynaptic CB_1 receptor, dictating whether exogenous phytocannabinoids produce homeostatic recalibration, dull sedation, or acute hyper-excitability.
Metabolic Clearance: Variations in cytochrome P450 enzymes—particularly CYP2C9 and CYP2C19—yield vast discrepancies in the half-life and active conversion of Delta9-THC into the significantly more potent 11-hydroxy-THC. A low-metabolizing individual can experience extended systemic saturation and subsequent receptor uncoupling from a dose that a rapid metabolizer clears without disruptive adaptation.
Degradatory Hydrolysis: Genomic variation in FAAH (such as the well-documented C385A variant) and MAGL changes how effectively an organism terminates endogenous anandamide and 2-AG signaling. What diagnostic manuals score as "tolerance" is frequently an individual's distinct baseline enzymatic set point interacting with exogenous inputs.
Downstream Monoaminergic Signaling: Variations in AKT1 and COMT dictate whether striatal and prefrontal circuits interpret cannabinoid agonism with adaptive plasticity or neurochemical friction.
To take millions of individuals exhibiting entirely distinct genomic pathways, enzymatic clearance rates, and neurochemical baselines and corral them into a single diagnostic diagnosis is biologically illiterate. It mistakes the exogenous botanical trigger for the condition itself, ignoring that the variance lies overwhelmingly in the patient's intrinsic biological wiring.
Therapeutic Impotence: When Diagnosis Fails to Deliver Treatment
The pragmatic justification for enduring a flawed diagnostic category is clinical utility: does the label unlock an effective, reproducible cure? For CUD, empirical evidence reveals a stark therapeutic vacuum.
After decades of dedicated clinical research:
There remains not a single FDA-approved pharmaceutical agent for CUD. Clinical trials investigating everything from synthetic partial agonists to FAAH inhibitors have repeatedly hit efficacy ceilings, precisely because they attempt to apply a single neurochemical lever to a heterogenous patient population that shares no common biological pathophysiology.
Standardized psychosocial modalities—Cognitive Behavioral Therapy and Motivational Enhancement Therapy—exhibit 12-month continuous abstinence rates ranging from 14% to 20%.
Even resource-intensive operant conditioning programs like Contingency Management, which pay patients cash vouchers for clean urinalysis, see dramatic relapse cascades the moment behavioral incentives are withdrawn.
When an intervention fails four out of five patients over the long arc, it does not demonstrate the intractability of an addiction; it reveals the invalidity of the clinical model. Standardizing treatment around an amorphous behavioral phenotype guarantees standardized therapeutic failure.
The Natural History Paradox: Spontaneous Remission Over Intervention
The fatal blow to CUD’s status as a distinct clinical pathology is its epidemiological trajectory outside the clinic. In classical pathology, an untreated chronic illness progresses toward systematic deterioration. Yet large-scale longitudinal surveys—such as the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC)—consistently establish that the vast majority of individuals who meet criteria for CUD resolve their symptoms without formal clinical intervention.
Over a multi-year timeline, roughly 50% to 65% of individuals undergo spontaneous remission. As environmental demands shift, life milestones arrive, and individuals organically identify when a particular consumption pattern no longer aligns with their cognitive or physical baseline, use is modified or abandoned entirely.
When spontaneous, self-guided recalibration vastly outperforms formal clinical paradigms, the moral and scientific justification for retaining the diagnostic construct vanishes. Categorizing dynamic lifestyle transitions, sub-optimal coping strategies, or uncalibrated self-medication as a chronic "substance use disorder" does not protect the patient; it creates an artificial pathology that medicalizes personal adaptation.
Toward Disaggregation and Precision Medicine
Eliminating Cannabis Use Disorder does not mean ignoring acute distress or physiological complications. Rather, it demands disaggregating this catch-all category into objective, mechanistically coherent realities:
Clear Physiological Conditions: True physical manifestations like Cannabinoid Hyperemesis Syndrome (CHS) should be treated as specific gastrointestinal and neuroendocrine toxicities with targeted clinical algorithms, rather than vague subsets of behavioral addiction.
Primary Psychiatric Etiology: When individuals consume cannabis excessively to manage untreated PTSD, ADHD, or atypical depressive architecture, clinical medicine must treat the primary neurological etiology directly, rather than masking it under a secondary "substance use" diagnosis that blames the patient's crude attempt at neurochemical self-regulation.
Pharmacogenomic Guidance: Replacing subjective behavioral checklists with biomarker and genomic profiling will allow clinicians to counsel individuals accurately based on their specific CYP2C9 clearance rates, CNR1 sensitivity, and affective risk markers.
The categorization of illness must serve understanding and healing. Cannabis Use Disorder has done neither.
By pathologizing individual genetic divergence and maintaining a taxonomy that fails both basic pharmacological logic and clinical trial rigor, the construct stands as a relic of 20th-century behavioral generalization. Retiring it is the first prerequisite for a truly biological, individualized approach to modern neurobiology and patient care.
When a diagnostic category fails to reflect biological reality and fails to produce therapeutic results, it does not merely sit harmlessly on a ledger—it distorts research priorities, misallocates public health capital, and assigns a lifelong clinical stigma to what is often nothing more than constitutional variance and developmental maturation.
We can no longer afford to preserve an administrative artifact out of institutional inertia. Every dollar spent standardizing generic behavioral treatments for an illusory monolith is a dollar stolen from investigating the true molecular architecture of the endocannabinoid system, identifying actionable pharmacogenomic biomarkers, and developing targeted interventions for genuine physiological toxicities.
The path forward demands decisive reform across three fronts:
For the American Psychiatric Association and Nosologists: Formally initiate the review process to retire Cannabis Use Disorder from the Diagnostic and Statistical Manual. Replace this blunt rubric with discrete, mechanistically defined physical diagnoses alongside rigorous secondary assessments of primary psychiatric disease.
For Clinical Researchers and Funding Bodies: Redirect grant allocations away from generalized psychosocial addiction protocols and toward precision pharmacogenomics—mapping how CYP2C9 polymorphisms, CNR1 variants, and metabolic baselines dictate clinical response.
For Clinicians and Care Providers: Discontinue the reliance on crude behavioral checklists. Screen patients for underlying trauma, affective disorders, and neurodivergence before labeling their attempts at neurochemical self-regulation as a primary pathology.
Medicine advances not only by what it discovers, but by what it is brave enough to discard.
It is time to retire Cannabis Use Disorder, dismantle the behavioral monolith, and build a psychiatry grounded in molecular reality, empirical epidemiology, and individual human biology.
