Cannabis can produce dramatically different emotional effects from one person to another. For some users, THC brings relaxation, laughter and relief from stress. For others, especially at higher doses or in uncomfortable surroundings, the same drug can trigger racing thoughts, paranoia, panic and an intense feeling that something is wrong. Scientists have long known that this paradox involves the brain’s cannabinoid system, but exactly how cannabinoids transform a stressful situation into an exaggerated fear response has remained difficult to explain.
A new study from Northwestern University provides one of the clearest biological explanations yet. Published in Nature Communications in October 2026, the research identified a specific population of neurons in the central amygdala that becomes unusually active after cannabinoid exposure, particularly when an animal encounters a threatening situation. The cells, known as somatostatin-expressing neurons, appear to help convert cannabinoid signaling into heightened avoidance and defensive behavior. When researchers selectively silenced those neurons, much of the cannabinoid-induced threat avoidance disappeared.
Scientists Focused on the Brain’s Fear and Stress Center
The amygdala has been central to anxiety research for decades. Located deep within the brain, this collection of interconnected structures helps evaluate danger, assign emotional significance to experiences and coordinate defensive responses. The central amygdala, or CeA, functions as an important output center within this system, influencing behavioral and physiological reactions to perceived threats. Previous research had already shown that cannabinoids can strongly alter activity within the amygdala, but scientists had not established which specific cell types were responsible for turning cannabinoid exposure into anxiety-like behavior.
Farhana Yasmin, Sachin Patel and colleagues at Northwestern University focused on central amygdala neurons that produce the signaling peptide somatostatin. These cells are involved in emotional learning, avoidance behavior and defensive responses. Rather than simply measuring whether the amygdala became more or less active after cannabinoid exposure, the team monitored individual neurons while mice encountered a naturally threatening stimulus. This allowed researchers to observe how cannabinoids changed the way specific brain-cell populations represented danger in real time.
Predator Odor Revealed the Cannabinoid-Fear Interaction
To create a threatening but controlled situation, researchers exposed mice to an odor derived from fox urine, a commonly used predator cue in neuroscience experiments. Mice naturally recognize predator-associated odors as dangerous and respond through behaviors such as cautious investigation, avoidance and freezing. Before encountering the odor, animals received either an inactive treatment or a cannabinoid receptor agonist. Researchers tested multiple doses to determine whether stronger cannabinoid exposure produced stronger defensive responses.
The results showed a clear dose-dependent pattern. Cannabinoid-treated mice spent less time investigating the predator odor and exhibited stronger defensive behaviors, including increased freezing. The combination of cannabinoid exposure and environmental threat appeared particularly important. The drug did not simply create one uniform behavioral state; instead, it amplified the animals’ responses when something threatening was present. That interaction closely resembles reports from human cannabis users who feel relaxed in one setting but suddenly become anxious when their environment becomes unpredictable, unfamiliar or stressful.
Tiny Microscopes Tracked Individual Neurons During Fear
The Northwestern team implanted miniature microscopes that allowed them to record calcium activity from individual somatostatin neurons in the central amygdala while the mice moved freely. Calcium activity is commonly used as an indicator of when neurons are active. This technique gave researchers a window into how specific cells responded before and after cannabinoid exposure and as animals approached or retreated from the threatening odor.
Cannabinoids substantially increased the spontaneous activity of these somatostatin neurons. The drugs also changed how groups of neurons organized themselves into functional patterns representing different threat-related behaviors and locations. Instead of simply raising activity uniformly across the amygdala, cannabinoid exposure appeared to reorganize the neural population so that certain groups of cells became more strongly associated with avoidance, threat investigation and behavioral state. The brain was effectively processing the threatening environment differently after cannabinoid receptors had been activated.
Somatostatin Neurons Were Necessary for Increased Threat Avoidance
Finding neurons that become active during anxiety does not necessarily mean those cells cause anxious behavior. To answer that question, the researchers genetically manipulated the somatostatin neurons so their activity could be suppressed. They then repeated the predator-odor experiments while the animals were under the effects of the cannabinoid.
Silencing the neurons significantly reduced the cannabinoid-induced increase in threat avoidance. The mice became more willing to investigate the threatening odor instead of keeping their distance. Interestingly, suppressing the cells did not eliminate every defensive response. Cannabinoid-enhanced freezing remained largely intact, suggesting that different aspects of fear may be controlled by separate neural pathways. Somatostatin neurons in the central amygdala therefore appear particularly important for avoidance behavior rather than acting as a single master switch controlling all cannabinoid-related fear.
Cannabinoids May Trigger Anxiety by Releasing a Neural “Brake”
Perhaps the most interesting part of the study involved explaining how cannabinoid receptor activation caused somatostatin neurons to become more active. At first glance, the result seems paradoxical because activation of cannabinoid type-1 receptors, or CB1 receptors, usually suppresses neurotransmitter release. THC and many other cannabinoids activate CB1 receptors located on nerve terminals, reducing the release of chemical messengers into the synapse.
Brain-slice experiments provided an explanation. Cannabinoid receptor activation reduced inhibitory GABA signaling onto the somatostatin neurons. GABA normally acts as one of the brain’s major braking systems, preventing neurons from becoming excessively active. By suppressing GABA release from nearby nerve terminals, cannabinoid signaling effectively removed part of that inhibition. The somatostatin neurons were therefore released from their normal restraint and became more active. A drug that suppresses neurotransmission at one point in a circuit can consequently increase activity further downstream by suppressing the neurons responsible for inhibition.
Higher Doses and Stress May Work Together
The findings help explain one of the best-known characteristics of THC: its emotional effects are strongly dose dependent. Small amounts of THC may produce relaxation or reduced anxiety in some people, while larger doses are considerably more likely to cause nervousness, paranoia or panic. Cannabinoid pharmacology does not operate through a single brain circuit, and different doses can affect networks involved in reward, memory, sensory processing and threat detection in different ways.
The Northwestern study suggests that environmental stress may amplify this dose effect. Strong cannabinoid signaling combined with a threatening stimulus produced particularly pronounced activation of the central amygdala system. Senior researcher Sachin Patel described the process as cannabinoids and environmental stress working together to release the inhibitory brake on the amygdala. In practical terms, the finding provides a possible neurological explanation for why someone can initially feel relaxed after using cannabis but become extremely anxious when confronted with an unexpected social interaction, frightening thought or unfamiliar environment.
Human Research Has Already Linked THC, Anxiety and the Amygdala
Although the new experiments were performed in mice, previous human research supports the broader idea that THC can alter brain systems involved in fear. A 2017 study published in Scientific Reports used functional MRI and positron emission tomography to study healthy volunteers who received 10 milligrams of oral THC or placebo. THC increased subjective anxiety and changed activity in the right amygdala while participants processed fearful stimuli. The effects were related to the availability of CB1 receptors, providing direct evidence in humans that cannabinoid signaling within fear-processing regions contributes to THC-induced anxiety.
Earlier neuroimaging research published in the Archives of General Psychiatry, now JAMA Psychiatry, also demonstrated that THC and cannabidiol can have very different emotional effects. THC increased anxiety, intoxication and psychotic-like symptoms, while CBD showed a tendency to reduce anxiety. The two cannabinoids produced different patterns of activation while participants viewed fearful faces, reinforcing the idea that “cannabis” cannot be treated as a pharmacologically uniform substance. Products dominated by THC may interact with emotional brain circuits differently from preparations containing substantial amounts of CBD.
Another 2026 Study Identified a Different Anxiety Circuit
The central amygdala is unlikely to be the only pathway explaining cannabinoid-induced anxiety. Another Nature Communications study published earlier in 2026 identified a separate circuit connecting the anterior cingulate cortex to the dorsomedial striatum. Researchers found that activating CB1 receptors reduced excitatory communication along this pathway in mice. When the scientists genetically removed CB1 receptors specifically from this circuit, animals displayed less innate anxiety-like behavior and reduced aversion to THC.
Together, the two studies illustrate the complexity of cannabinoid signaling. THC affects CB1 receptors across much of the brain, and different circuits may contribute to different aspects of anxiety. Suppressing one cortical pathway can promote avoidance, while reducing inhibition inside the amygdala can activate threat-responsive neurons. The emotional result experienced by a cannabis user may therefore reflect the combined activity of multiple networks rather than a single “anxiety center.”
Why THC Can Feel Relaxing and Anxiety-Producing at the Same Time
The human endocannabinoid system normally helps regulate stress responses by releasing internally produced cannabinoids such as anandamide and 2-arachidonoylglycerol when and where they are needed. These molecules typically act locally and temporarily, fine-tuning communication between neurons. THC differs because consuming cannabis can activate CB1 receptors across many brain regions simultaneously and for a much longer period than normal endocannabinoid signals.
That helps explain why THC can produce opposite effects depending on dose, context and individual biology. Cannabinoid signaling can reduce stress activity in some circuits while increasing threat sensitivity in others. Previous animal studies have shown that enhancing certain endocannabinoids can produce anti-anxiety effects, yet direct CB1 receptor stimulation at high levels can become anxiogenic. The nervous system appears to depend on carefully balanced cannabinoid signaling; overwhelming that system with a potent external cannabinoid can shift circuits away from their normal equilibrium.
Potency May Be Increasing the Importance of Cannabis Anxiety
The question has become more relevant as the potency of commercial cannabis has increased. Marijuana products available several decades ago typically contained much lower THC concentrations than many modern flower varieties, while concentrates and vapor products can deliver far higher doses. Edibles also create a unique risk because their effects are delayed, making it easier for inexperienced consumers to take additional THC before the first dose becomes fully active.
Emergency departments regularly treat patients experiencing acute cannabis-related anxiety, panic, vomiting, confusion and rapid heart rate. Severe reactions are rarely medically life-threatening from THC alone, but they can be extremely frightening and may require observation or treatment. The new neuroscience suggests that these experiences are not simply the result of a user “thinking negatively.” Cannabinoid exposure can physically alter how threat-processing networks operate, making ambiguous or stressful situations feel more dangerous.
Individual Biology Still Matters
Not everyone reacts to THC in the same way. Genetics, previous cannabis exposure, tolerance, age, baseline anxiety, psychiatric history, hormone levels and other biological factors can all influence the response. Experienced users may develop tolerance to certain THC effects, while people who are highly sensitive to anxiety can experience panic at doses that others find comfortable. Cannabis products also differ enormously in their cannabinoid and terpene composition.
CBD is particularly important because several studies suggest it can counteract some effects of THC under certain conditions, although results vary depending on dose and timing. Research involving adolescent animals and primates has also found that CBD can reduce some THC-related changes in the amygdala, including inflammatory responses in specific experimental models. These findings do not mean adding CBD automatically prevents anxiety, but they reinforce the idea that cannabinoid composition matters when predicting how a cannabis product will affect the brain.
Mouse Findings Need Confirmation in Humans
The Northwestern researchers used a synthetic cannabinoid agonist rather than having mice consume conventional marijuana flower. Synthetic experimental agonists can activate cannabinoid receptors differently and sometimes more strongly than THC, making it important not to treat the mouse results as a direct measurement of what happens in every person who uses cannabis. Predator odor is also an experimental model of threat and cannot reproduce the complexity of human panic, paranoia or social anxiety.
Even with those limitations, the cellular mechanism is compelling because it connects several pieces of existing evidence. Cannabinoids activate CB1 receptors in the central amygdala, stress and cannabinoids have previously been shown to interact in this brain region, somatostatin neurons participate in defensive behavior, and human imaging studies already link THC-induced anxiety with altered fear-processing networks. Future studies will need to determine whether comparable somatostatin neuron activity can be detected in humans and whether differences in these circuits explain why some people are unusually susceptible to cannabis-induced anxiety.
The Discovery Could Eventually Point Toward New Anxiety Treatments
The study may ultimately have implications beyond medical marijuana. Somatostatin neurons in the central amygdala are part of the brain’s broader threat-response system, meaning the same cells could contribute to pathological anxiety even without cannabinoid exposure. If excessive activity in this neuronal population represents a final common pathway for heightened avoidance, researchers may eventually be able to target the system when developing new treatments for anxiety disorders.
For cannabis science, the discovery provides a more precise explanation for one of THC’s most puzzling effects. Cannabis does not create anxiety simply because it is stimulating or because users expect to become paranoid. Cannabinoid receptor activation can change the balance between inhibitory and excitatory signaling inside one of the brain’s most important fear centers. Under stressful conditions and at stronger doses, that shift can release threat-responsive neurons from inhibition and amplify defensive behavior. The finding moves researchers closer to understanding why cannabis can feel calming one moment and frightening the next—and why dose, environment and individual brain biology can determine which experience ultimately wins.
