What are marijuana's effects?

When cannabis researchers talk about the endocannabinoid system, they sometimes call it the body’s own cannabis-like network — which is backwards. Cannabis doesn’t mimic the body. The body built a system that THC happens to be able to fool.

The endocannabinoid system (eCB) predates cannabis by hundreds of millions of years. It’s in vertebrates, in sea squirts, arguably in some invertebrates. Its two primary receptors — CB1 and CB2 — are among the most abundant receptor types in the human brain. CB1 is concentrated in areas that regulate memory, appetite, pain, mood, and the executive functions of the prefrontal cortex. CB2 is found throughout the immune system. These receptors exist to respond to the body’s own endocannabinoids — molecules like anandamide (from the Sanskrit word ananda, meaning bliss) and 2-arachidonoylglycerol, or 2-AG. They are manufactured on demand, used locally, and rapidly metabolized. They serve as retrograde messengers, allowing neurons to talk back to the cells signaling them — a feedback system for modulating everything from pain responses to reward signals to anxiety.

THC is not that. THC is a partial agonist at CB1 that doesn’t break down the way the body’s own endocannabinoids do. It stays. It spreads. And the effects of a molecule that binds to the most abundant receptor in the brain, stays there longer than it should, and does so in concentrations the brain was never designed to encounter — those effects are not the effects of bliss. They are the effects of a system being run past its design parameters.

Why THC Is Not the Same as Anandamide, and Why That Matters for Dependence

The difference between anandamide and THC isn’t just potency. It’s duration and distribution.

Anandamide is synthesized and released locally, in response to specific synaptic conditions, and is degraded rapidly by the enzyme FAAH. Its action is targeted and brief — a signal, not a state. THC is inhaled or ingested, distributed through the bloodstream to all tissues expressing CB1, and metabolized slowly. It can remain in fat tissue for weeks. Its effects on CB1 aren’t the body’s own precise signals; they’re more like a sustained broadcast across every channel simultaneously.

The consequence, with repeated exposure, is predictable: downregulation. When CB1 receptors are bombarded with a partial agonist over days and weeks, the brain does what it does with any overstimulated system — it reduces receptor density and sensitivity. This is the biological mechanism of tolerance. More THC is required to produce the same effect. And when THC is withdrawn — when the external agonist is removed — the endocannabinoid system is suddenly underactive, because the receptors have been downregulated in response to the sustained external stimulation. This produces withdrawal: irritability, sleep disruption, anxiety, appetite changes, restlessness. These symptoms are real. They are not as severe as opioid or benzodiazepine withdrawal, and they are not life-threatening. But they are biological, and they are the mechanism by which cannabis use disorder maintains itself.

Cannabis use disorder (CUD) affects approximately 9–10% of people who ever use cannabis, rising to about 17% of those who begin using in adolescence, when the developing prefrontal cortex and limbic system are particularly vulnerable to CB1 modulation. Those numbers are not alarming if the average exposure is occasional weekend use of low-potency flower. They become more alarming when the exposure is daily use of high-potency concentrates.

The Discontinuity of High-Potency Products

Here is the trend that the eCB system’s biology makes clinically significant.

In the 1970s, the average THC concentration in cannabis confiscated by law enforcement was 1–2%. By the 1990s, it had risen to 4–7%. Today, flower sold in licensed commercial markets commonly tests at 20–30% THC. Concentrates — dabs, shatter, wax, live resin, distillate — routinely reach 70–90%. This is not the same drug at higher dose. This is a quantitative change large enough to produce qualitative differences in effect, dependence risk, and adverse event profile.

Concentrates — dabs, shatter, wax, live resin, distillate — routinely reach 70–90%.

A 2026 international study from the University of Bath, published in The Lancet Psychiatry this month, found something critical about the relationship between legal market design and these potency trends: decriminalization — removing criminal penalties for possession while not creating commercial markets — does not increase cannabis use or addiction rates. But commercial legalization with for-profit markets does. The Bath team’s analysis of data from the U.S., Canada, Europe, Australia, and New Zealand found that commercial markets are associated with increased product potency and higher rates of cannabis use disorder, driven by market incentives that favor high-margin, high-THC products over lower-potency alternatives.

This is a policy and biology story at the same time. The biology predicts that higher-potency products produce faster receptor downregulation, faster tolerance, and faster CUD onset. The policy finding confirms that markets designed for profit produce higher-potency products. The implication is that recreational legalization frameworks that don’t include explicit potency regulation are inadvertently optimizing for addiction risk.

What the 54-Study Meta-Analysis Actually Found

In March 2026, The Lancet Psychiatry published a systematic review and meta-analysis of 54 randomized controlled trials examining cannabinoids as treatments for mental health conditions and substance use disorders. The review covered 2,477 participants across studies conducted between 1980 and May 2025.

The findings were largely negative. For anxiety disorders, PTSD, psychotic disorders, OCD, and anorexia nervosa, cannabinoids showed no statistically significant therapeutic benefit over placebo. For opioid use disorder, no benefit. The review found evidence of benefit for two conditions: insomnia and Tourette’s syndrome.

The finding that got the most clinical attention: cannabinoids significantly increased cocaine craving in study participants with cocaine use disorder. This is not what advocates for cannabis as a harm reduction tool for stimulant use expected, and it aligns poorly with the narrative that cannabis can substitute for more dangerous drugs. The biological mechanism here isn’t fully understood, but CB1 modulation in the mesolimbic dopamine system — the reward circuit — interacts with the same circuitry that cocaine dysregulates. Stimulating it further does not appear to help people trying to reduce cocaine use.

The review has limits: heterogeneous study designs, varying cannabinoid formulations, small sample sizes in individual trials. It does not settle every question about medicinal cannabinoids. CBD, which does not strongly bind CB1 or CB2 and works through distinct mechanisms including TRPV1 channels, was underrepresented in the RCT evidence base and may have different therapeutic profiles — particularly for anxiety and seizure disorders, where regulatory approval (Epidiolex for severe pediatric epilepsy) already reflects a genuine evidence base.

But the meta-analysis does settle one question: the casual claim that cannabis, broadly defined, is an effective treatment for addiction or most mental health conditions is not supported by the controlled trial evidence. The pharmacology predicted this. CB1 agonism during active mental health conditions is unlikely to produce consistent benefit when those conditions involve exactly the circuits CB1 modulates — reward, anxiety, cognition, mood.

What This Means for Clinical Practice

Cannabis use disorder is underdiagnosed and undertreated. There is no FDA-approved pharmacotherapy for CUD — no medication equivalent to buprenorphine for OUD or naltrexone for alcohol. Behavioral interventions, particularly cognitive-behavioral therapy and contingency management, have demonstrated efficacy in reducing use and managing withdrawal. But the treatment infrastructure for CUD is thin relative to prevalence.

As commercial cannabis markets mature and high-potency products become the default in legalized states, providers should expect to see more patients presenting with CUD features — tolerance, withdrawal, use despite consequences — who don’t identify as having a “drug problem” because cannabis is legal and normalized. The biology doesn’t care whether a substance is legal. Receptor downregulation doesn’t check the statutory schedule.

The endocannabinoid system is one of the most fundamental regulatory networks in the human body. It governs aspects of pain, appetite, memory consolidation, emotional regulation, and immune function. Understanding how THC disrupts it isn’t an argument against cannabis. It’s an argument for approaching cannabis — like every other pharmacologically active substance — with respect for the biology, not the marketing.

It’s an argument for approaching cannabis — like every other pharmacologically active substance — with respect for the biology, not the marketing.

The marketing says cannabis is medicine. Sometimes it is — Epidiolex is real, Marinol is real, the evidence for certain pain applications is real. What the marketing doesn’t say is that the same plant, at the same time, in high concentrations, activates the same biological system in ways that create dependence in roughly one in ten users. Both things can be true. The biology requires that they be.