Last verified: April 2026

Glaucoma — Why Ophthalmologists Say No

Glaucoma was among the first medical conditions associated with cannabis, based on a seminal finding by Robert Hepler and Ira Frank at UCLA, published in JAMA in 1971. They demonstrated that smoking cannabis reduced intraocular pressure (IOP) by 25–30% in both normal subjects and glaucoma patients. This finding was real, replicable, and mechanistically sound — cannabinoid receptors are present in the ciliary body and trabecular meshwork, where aqueous humor production and drainage are regulated.

The problem is entirely practical. The IOP-lowering effect lasts only 3–4 hours, meaning effective glaucoma management with cannabis would require 6–8 doses per day around the clock. Glaucoma is a chronic, progressive disease requiring consistent 24-hour IOP control to prevent optic nerve damage and irreversible vision loss. No patient can realistically maintain that dosing frequency, and the psychoactive effects at the required cumulative daily dose would be severely impairing.

Moreover, a 2018 study by Sally Miller and colleagues at Indiana University, published in Investigative Ophthalmology & Visual Science, demonstrated that CBD may actually increase IOP — the opposite of the desired therapeutic effect. In a mouse model, topical CBD significantly raised IOP, potentially counteracting any benefit from co-administered THC. This finding is particularly concerning given the prevalence of high-CBD, low-THC products marketed to older adults — the demographic most likely to have glaucoma.

The position of every major ophthalmology organization is unambiguous. The American Academy of Ophthalmology (AAO), the American Glaucoma Society (AGS), and the Glaucoma Research Foundation (GRF) all recommend against cannabis for glaucoma. Modern glaucoma treatment uses prostaglandin analogues (latanoprost), beta-blockers (timolol), or surgical interventions that provide 24-hour IOP reduction with once- or twice-daily dosing and no psychoactive effects. Cannabis is not merely inferior — it is impractical as a glaucoma treatment by any clinical standard.

The American Academy of Ophthalmology does not recommend marijuana or other cannabis products for the treatment of glaucoma. The effect is short-lived, and there is no evidence that marijuana use can alter the course of the disease.

American Academy of Ophthalmology, Position Statement 2014 (reaffirmed 2024)

Fibromyalgia and Migraine — The CED Hypothesis

Fibromyalgia and migraine are frequently cited as conditions that respond to cannabis, and both are qualifying conditions in multiple medical cannabis states. The theoretical framework supporting these claims is the Clinical Endocannabinoid Deficiency (CED) hypothesis, proposed by neurologist Ethan Russo and published in Neuroendocrinology Letters in 2004, with an updated review in 2016.

Russo’s hypothesis proposes that fibromyalgia, migraine, and irritable bowel syndrome (IBS) share a common underlying pathophysiology: deficient endocannabinoid system tone. The theory posits that reduced levels of anandamide and 2-AG, decreased CB1 receptor expression, or enhanced FAAH/MAGL enzymatic degradation of endocannabinoids creates a systemic deficiency that manifests as the overlapping symptom complexes seen in these conditions — diffuse pain, central sensitization, gastrointestinal dysfunction, and sleep disturbance.

The CED hypothesis is intellectually coherent and biologically plausible. Endocannabinoid system measurements in fibromyalgia patients have shown some differences compared to controls, and the symptom overlap between CED-candidate conditions is consistent with a shared underlying mechanism. However, the hypothesis has not been validated by prospective clinical trials. No study has demonstrated that fibromyalgia or migraine patients have consistently measurable endocannabinoid deficits, and no randomized trial has shown that exogenous cannabinoid supplementation corrects such a deficit and improves clinical outcomes.

The available evidence consists of surveys and observational studies showing that fibromyalgia and migraine patients who use cannabis report symptom improvement. These data are subject to the full complement of observational biases: self-selection, placebo response, recall bias, and confounding by concurrent treatments. They are consistent with the CED hypothesis but do not prove it — they are equally consistent with cannabis providing nonspecific symptomatic relief (analgesic, anxiolytic, sedative) without addressing any underlying endocannabinoid deficit.

Parkinson’s Disease — The Evidence Gap

Parkinson’s disease (PD) is a qualifying condition in several medical cannabis states, despite an evidence base that is thin and largely negative for the motor symptoms that define the disease. The theoretical rationale involves the dense CB1 receptor expression in the basal ganglia — the brain region most affected by PD — and preclinical evidence suggesting that cannabinoids could modulate dopaminergic and non-dopaminergic circuits involved in motor control.

The most directly relevant clinical trial was conducted by Christopher Carroll and colleagues, published in Neurology in 2004. This randomized, double-blind, placebo-controlled crossover study tested oral cannabis extract (containing 2.5 mg THC per capsule) in PD patients with levodopa-induced dyskinesia — the involuntary movements that develop as a complication of long-term dopamine replacement therapy. The primary outcome was negative: cannabis extract did not significantly reduce dyskinesia compared to placebo on any objective or subjective measure.

Observational studies and surveys of PD patients using cannabis report improvements in non-motor symptoms — sleep, pain, anxiety, and quality of life — rather than the cardinal motor features (tremor, rigidity, bradykinesia). This pattern suggests that cannabis may provide nonspecific symptomatic relief rather than targeting the dopaminergic pathology underlying PD. Until larger, longer trials specifically test cannabis in defined PD symptom domains, the evidence remains insufficient by NASEM criteria.

Inflammatory Bowel Disease — The CBG Angle

Inflammatory bowel disease (IBD), encompassing Crohn’s disease and ulcerative colitis, is another condition where patient self-report is enthusiastic but controlled trial data are limited. The most significant preclinical finding involves not THC or CBD but cannabigerol (CBG).

Borrelli et al. (2013), published in Biochemical Pharmacology, demonstrated that CBG reduced inflammation in a murine model of colitis, attenuating nitric oxide production, reducing myeloperoxidase activity in the colon, and decreasing IL-1β, IL-10, and interferon-γ levels. The effect was comparable to the reference anti-inflammatory agent sulfasalazine. This preclinical work provides a mechanistic rationale for the “CBG for gut health” claims that have entered the consumer market — but the gap between a mouse colitis model and human IBD therapy is vast, and no human clinical trial of CBG for IBD has been completed.

For THC and CBD specifically, the most notable trial was Naftali et al. (2021), which tested CBD-rich cannabis in Crohn’s disease. While patients reported improved quality of life, endoscopic and inflammatory markers did not improve — a pattern reminiscent of the Sativex-Ashworth dissociation in MS spasticity, where subjective and objective outcomes diverge.

Autism Spectrum Disorder — The Aran Trial

Autism spectrum disorder (ASD) has become an increasingly common qualifying condition in medical cannabis programs, driven primarily by parental demand and observational data from Israeli clinics. The most rigorous data come from a randomized controlled trial led by Adi Aran at Shaare Zedek Medical Center in Jerusalem, published in Molecular Autism in 2021.

The trial enrolled 150 children and adolescents with ASD and significant behavioral problems. It used a three-arm design comparing whole-plant cannabis extract (20:1 CBD:THC ratio, containing the full spectrum of minor cannabinoids and terpenes), purified CBD and THC in the same 20:1 ratio, and placebo. The primary outcome was change in disruptive behavior as measured by the Home Situation Questionnaire — Autism Spectrum Disorder (HSQ-ASD).

The result was split. The whole-plant extract arm showed statistically significant improvement in disruptive behaviors compared to placebo. The purified CBD:THC arm did not significantly differ from placebo. This finding is provocative on multiple levels: it suggests that the combination of CBD, THC, and the additional minor cannabinoids and terpenes present in whole-plant extract may be necessary for therapeutic effect in ASD — a result consistent with the entourage effect hypothesis but requiring independent replication.

The mechanism by which cannabinoids might improve ASD-associated behavioral symptoms is speculative. Hypotheses include modulation of the oxytocin system (endocannabinoids facilitate oxytocin release in the nucleus accumbens), reduction of anxiety (a major driver of behavioral outbursts in ASD), and anti-inflammatory effects in a subset of ASD patients with elevated neuroinflammatory markers. None of these mechanisms have been confirmed in the ASD population specifically.

Alzheimer’s Disease — The Preclinical Promise

Alzheimer’s disease (AD) represents perhaps the largest gap between preclinical excitement and clinical evidence in cannabinoid research. The most publicized findings came from the Salk Institute for Biological Studies, where researchers demonstrated in cell culture that THC and other cannabinoids could reduce amyloid-beta accumulation and associated neuroinflammation. The endocannabinoid system is altered in AD brains, with changes in CB1 and CB2 receptor expression, FAAH activity, and endocannabinoid levels that suggest a potential therapeutic target.

Preclinical models have shown cannabinoid-mediated reductions in amyloid plaque burden, tau hyperphosphorylation, and microglial activation. CB2 agonists in particular have shown anti-neuroinflammatory effects in AD mouse models, and PPARγ activation by CBD and THC may provide neuroprotective effects through anti-inflammatory gene regulation.

However, the clinical gap is total. No randomized controlled trial has demonstrated that any cannabinoid slows cognitive decline, reduces dementia progression, or improves cognitive function in Alzheimer’s patients. The handful of small clinical studies have tested cannabinoids for behavioral symptoms of dementia (agitation, aggression, sleep disturbance) rather than disease modification, with mixed results. Dronabinol showed modest improvement in nighttime agitation in one small trial, but the data are insufficient to support clinical recommendations.

The Alzheimer’s evidence perfectly illustrates a recurring theme in cannabinoid research: the endocannabinoid system is clearly involved in the pathophysiology, the preclinical data are encouraging, but the translation from laboratory to clinic has not occurred. Until it does, claims that cannabis “treats” or “prevents” Alzheimer’s disease are not supported by human evidence.