What marijuana do the brain?

The short version: cannabis is not a benign brain drug. It disrupts memory circuits, and heavy early use is tied to measurable cognitive and psychiatric harm.

The brain runs on a balance of excitation and inhibition. THC, the main psychoactive compound, tips that balance. It activates receptors on support cells in the hippocampus, the memory center, and triggers a flood of glutamate, the brain's main excitatory signal [3]. That flood weakens the connections that store memories, a process called long-term depression [3]. Repeated exposure inflames the same region and rewires gene expression, and in adolescent animals those changes outlast the exposure [3]. Heavy users show lower glutamate markers in several brain regions, though the studies are small and inconsistent [1].

The human evidence is real but mostly observational. A review of imaging studies found consistent structural changes in the hippocampus, amygdala, and frontal regions of chronic users [4]. The most striking human data come from a small group of very heavy, isolated users with no other drug use: they scored worse on a broad cognitive composite and showed schizotypal traits, with the biggest gap on a visuospatial task [2]. A large birth cohort found persistent use linked to broad cognitive decline, worst in adolescent-onset users, and stopping did not fully reverse it [5]. Twin data tie cannabis to more psychotic-like experiences, though the brain-network measure did not explain the link [6].

What this retrieval cannot see: it holds no randomized trial of cannabis on brain outcomes, no long-term follow-up past the cohort above, and almost no female data, since the mechanistic studies were mostly in male animals [3]. The harms are best documented in heavy, early, daily users. Light occasional adult use is far less studied, and the evidence there is thin. The psychiatric risk is not evenly distributed: people with certain genetic variants face much higher odds of cannabis-induced psychosis [5].

My call: for a developing brain, the documented memory and psychiatric risks outweigh any claimed benefit, and the case against heavy adolescent use is strong. Confidence: moderate.