JAMA, Medical Cannabis and the Endocannabinoid Blind Spot

There is a saying in science: Absence of evidence is not evidence of absence. But in the case of the recent JAMA review on the therapeutic use of cannabis, I would propose a correction:

 Absence of mechanistic insight leads to the illusion of insufficient evidence.

A new medical cannabis review published this week in JAMA by Hsu, Shah, and Jordan concludes that there is “insufficient evidence” for the use of cannabis in treating most medical conditions, including insomnia and acute pain. They warn of risks, cite modest effect sizes, and ultimately advise caution.

From the standpoint of Addiction Psychiatry, the specialty of the review’s authors, this conclusion is logical. They are trained to view psychoactive compounds through the lens of harm reduction, dependency, and abuse potential. When you look at cannabis solely as a “drug” to be audited against a symptom checklist, the results are indeed messy.

But from the standpoint of Endocannabinoid System (ECS) Biology, this review is a frustrating example of looking at the smoke while ignoring the fire.

The “Drug” vs. The “System”

The fundamental flaw in this review (and many of the randomized trials it analyzes) is a category error. Cannabinoids are treated like allopathic pharmaceuticals—single molecules designed to suppress a specific pathway (for example, blocking a sodium channel to stop pain).

The question being asked is:

“Does cannabis treat chronic pain?”

The data comes back mixed.

Conclusion: “Insufficient evidence.”

This is the wrong question. The question we ask at ECS.education is:

“Is the patient’s pain a symptom of Endocannabinoid System Dysfunction, and can we restore homeostasis?“

If you administer insulin to a healthy person, it causes harm. If you administer it to a person with diabetes, it can save their life. The difference is not the drug; it is the metabolic state of the patient. By lumping “everyone with pain” or “everyone with insomnia” into large cohorts without assessing baseline ECS tone, these studies dilute the data. Patients with clinical endocannabinoid deficiency (CED) are mixed with those who have normal or even excessive ECS signaling. The result is statistical noise, not biological truth.

NCDs are ECS Failure States

The JAMA authors claim insufficient evidence for cannabis in treating non‑communicable diseases (NCDs) such as chronic pain, metabolic disorders, and neurodegenerative or organ‑specific pathologies. This stands in stark contrast to the 2024–2025 literature, which increasingly frames these conditions as manifestations of ECS dysfunction rather than isolated symptom clusters.

Recent work on fibromyalgia, for example, explicitly validates an ECS‑centric etiology. García‑Domínguez describes how altered endocannabinoid signaling contributes to central sensitization and the characteristic pain amplification seen in fibromyalgia, positioning the ECS as a causal driver rather than a downstream modifier (García‑Domínguez, 2025).

In the neurological domain, Hakami and Alshehri systematically review clinical trials of cannabinoids in disorders such as multiple sclerosis, epilepsy, and movement disorders, emphasizing that many observed benefits align with restoration of disrupted ECS‑mediated homeostasis rather than simple symptomatic suppression (Hakami & Alshehri, 2025). Simankowicz and Stępniewska further broaden this picture by detailing how endocannabinoids participate in key physiological processes—metabolic regulation, immune modulation, stress responsivity—and how dysregulation of these pathways underpins diverse NCDs, from metabolic syndrome to chronic inflammatory states (Simankowicz & Stępniewska, 2025).

The same etiological theme is now being traced across organ systems. Salum and colleagues outline how aberrant endocannabinoid signaling influences tumor microenvironments, proliferation, and apoptosis, arguing that cancer biology itself must increasingly be understood through an ECS lens (Salum et al., 2025). Feki and co‑authors map ECS roles in ocular physiology and glaucoma, showing that endocannabinoid signaling is integral to intraocular pressure control and neuroprotection in the eye (Feki et al., 2025). Most integratively, Șerban and colleagues describe the ECS as a unifying molecular signaling network in human disease, integrating receptor pharmacology and therapeutic innovation and explicitly proposing ECS dysfunction as a common mechanistic substrate across multiple NCD categories (Șerban et al., 2025).

Against this backdrop, declaring there is “insufficient evidence” for therapeutic modulation of the ECS in NCDs is a bit like evaluating insulin while refusing to acknowledge the existence of diabetes: the molecules are being judged in isolation from the system they are meant to repair.

The Substrate Approach: Beyond Crude Agonism

There is one place where we agree with the JAMA authors: their warning against high‑potency cannabinoids. The concern is not misguided, but it is incomplete. High‑affinity receptor activation is a crude, blunt‑force intervention that ignores the molecular substrate in which these receptors operate.

The future of ECS therapeutics is not about hitting receptors harder. It is about substrate‑driven restoration, ensuring that the biological hardware can properly transduce the signal.

The ECS operates within the lipid bilayer.The fluidity and composition of this membrane—the balance between saturated fats, omega‑6 polyunsaturates, and the longer‑chain monounsaturated and omega‑3 species—fundamentally determines how well CB1 and CB2 receptors can function. A membrane composed predominantly of rigid, saturated lipids is a membrane that cannot support proper GPCR signaling, regardless of ligand availability. The same membrane rigidity and inflammatory lipid profile that characterize Western‑diet metabolic syndrome will also impair ECS signaling.

Rational ECS support therefore focuses on ECS substrates first, restoring the lipid environment in which these receptors are embedded before considering how aggressively to modulate the receptors themselves.

The Measurement Problem: CB1 as a Window into Brain Function

Here is something the JAMA review misses entirely: CB1 receptors are the most prevalent G protein–coupled receptors in the brain, present at densities far exceeding dopamine, serotonin, or many glutamate receptors. This is not an accident of evolution. CB1’s ubiquitous distribution and unique pharmacology make it a master regulatory hub for nearly every neurological function—from sleep architecture to pain processing to mood regulation.

Yet almost none of the studies the JAMA authors reviewed actually measure CB1 function directly. They rely on patient self-report (“Did your pain improve?”) or crude behavioral outcomes.

This is backwards. If we want to understand whether ECS modulation works, we need to develop and use direct measurement approaches that take advantage of CB1’s unique position as a window into whole‑brain physiology. The tools exist—ranging from non‑invasive neurophysiology to emerging receptor‑sensitive imaging—but the will and knowledge to apply them systematically is lacking.

Contrast this with how dopamine is studied: motor output is quantified, reward‑seeking is measured, and dopaminergic neurons can be imaged. Yet for the most abundant GPCR system in the brain, we continue to rely on patient questionnaires designed decades ago.

Conclusion

The JAMA review is, in many ways, an accurate audit of badly designed trials. It is a report on the limitations of the current medical paradigm, which tries to force complex botanical keys into reductive pharmacological locks.

They see “Insufficient Evidence.”
What we see is insufficient substrate and insufficient measurement.

As we move forward, the questions must change. Instead of asking “Does cannabis work?” we should be asking:

1. Does this patient have an underlying ECS dysfunction?
2. Is the biological substrate (lipid composition, membrane dynamics, receptor density) optimized for therapeutic intervention?
3. Can we measure the actual impact on CB1 function using approaches that reflect its unique position in brain physiology?

The answer lies not in higher doses of phytocannabinoids, but in the precise, molecular support of the biological machinery itself, and our commitment to measure what actually matters.

References:

Hsu M, Shah A, Jordan A, Gold MS, Hill KP. Therapeutic Use of Cannabis and Cannabinoids: A Review. JAMA. Published online November 26, 2025. doi:10.1001/jama.2025.19433

García-Domínguez M. Role of the Endocannabinoid System in Fibromyalgia. Curr Issues Mol Biol. 2025;47(4):230. Published 2025 Mar 27. doi:10.3390/cimb47040230

Hakami OM, Alshehri FS. Therapeutic potential of cannabinoids in neurological conditions: a systematic review of clinical trials. Front Pharmacol. 2025;16:1521792.

Simankowicz P, Stępniewska J. The Role of Endocannabinoids in Physiological Processes and Disease Pathology: A Comprehensive Review. J Clin Med. 2025;14(8):2851. Published 2025 Apr 21. doi:10.3390/jcm14082851

Salum KCR, Miranda GBA, Dias AL, Carneiro JRI, Bozza PT, da Fonseca ACP and Silva T (2025) The endocannabinoid system in cancer biology: a mini-review of mechanisms and therapeutic potential. Oncol. Rev. 19:1573797. doi: 10.3389/or.2025.1573797

Feki O, Zhioua-Braham I, Haddar S, Arfaoui A, Errais K, Feki M. The Endocannabinoid System: Role in Ocular Physiology and Therapeutic Potential in Eye Diseases: A Narrative Review. J Ocul Pharmacol Ther. 2025;41(9):503-517. doi:10.1177/10807683251368650

Șerban M, Toader C, Covache-Busuioc RA. The Endocannabinoid System in Human Disease: Molecular Signaling, Receptor Pharmacology, and Therapeutic Innovation. Int J Mol Sci. 2025;26(22):11132. Published 2025 Nov 18. doi:10.3390/ijms262211132