Scientists identify the nose-to-brain circuit that makes slow breathing calming

Slow nasal breathing patterns directly reduce anxiety by activating a specific pathway from the nose to the brain’s emotion centers, while fast breathing tends to have the opposite effect. The research also suggests that a daily routine of simulated slow breathing can provide lasting relief from anxiety-like symptoms. The findings were published in the Proceedings of the National Academy of Sciences (PNAS).

Practices like yoga and meditation often focus on slow, rhythmic breathing to calm the mind. While scientists know that the lower brainstem controls our automatic breathing rhythm, the physical sensation of air moving through the nasal cavity also sends powerful signals to the rest of the brain. When we inhale, the physical pressure of the air activates olfactory sensory neurons. These are specialized nerve cells located in the nasal cavity that detect both chemical odors and the mechanical pressure of airflow.

Past work indicates that these nasal signals play a role in shaping how we feel. For example, a study covered by PsyPost in 2016 provided evidence that the route and rhythm of breathing directly modulate brain activity, with nasal inhalation leading to faster recognition of fearful faces. However, it was not entirely understood exactly how the speed of these nasal signals affects emotional states, or which exact brain pathways translate the rhythm of breathing into a feeling of calm or panic.

“The idea that slow breathing calms us down is ancient, but the neuroscience was unclear,” Ruiqi Wu, a professor and principal investigator at the Institutes of Brain Science at Fudan University, told PsyPost. “It didn’t tell whether breathing affects emotion through the brain’s top-down processes, through brainstem respiratory centers, or through the airflow itself.”

“We wanted to settle this causally,” Wu continued. “Our answer: the airflow itself is the signal. Its frequency alone, independent of breathing effort or conscious control, shifts anxiety up or down.”

To test this, the research team, led by Xinsong Guo and supervised by Wu, designed a series of experiments to map this connection. They began by measuring the natural breathing rates of mice, establishing two breaths per second as a slow frequency, four breaths per second as a moderate frequency, and seven breaths per second as a fast frequency.

To manipulate the nasal signals, the researchers used a technique called optogenetics. They genetically modified the olfactory sensory neurons in the mice to be sensitive to light. By shining a blue laser through tiny fiber optic cables implanted in the animals’ nasal cavities, the scientists could control exactly how fast these neurons fired, without changing the animals’ actual breathing rate.

During the light stimulation, the mice were placed in behavioral arenas to assess their anxiety levels. In the Elevated Plus Maze, a mouse explores a raised cross-shaped platform with two enclosed arms and two open arms. In the Open Field Test, a mouse is placed in a large, empty box. Mice are prey animals that naturally fear open spaces where they might be seen, so they usually hug the walls or hide in the enclosed arms. If a mouse spends more time exploring the open areas, it indicates a lower state of anxiety.

The behavioral tests showed distinct changes depending on the speed of the stimulation. When the researchers stimulated the nasal neurons at a slow rate of two times per second, the mice spent much more time in the open areas of the testing arenas, indicating a calmer state. Fast stimulation at seven times per second caused the mice to avoid the open areas, suggesting heightened anxiety. Moderate stimulation produced no noticeable changes compared to normal behavior.

“How you breathe through your nose is not just a reflection of your emotional state; it is an input your brain actively reads,” Wu said. “Slow nasal breathing engages a specific brain circuit that lowers anxiety, while fast nasal breathing pushes the same circuit in the opposite direction. This gives scientific grounding to breathing practices, with a caution: rapid breathing during stress is not just a symptom of anxiety, it may actively feed it.”

To understand what was happening inside the brain, the team recorded electrical activity in a region called the perirhinal cortex. The perirhinal cortex is an area located deep within the brain that is often associated with memory but receives direct input from the olfactory system. The researchers focused on high-gamma waves, which are fast electrical rhythms associated with active neural processing and communication.

They found that slow nasal stimulation boosted high-gamma waves in the perirhinal cortex, while fast stimulation reduced them. A higher level of high-gamma activity correlated directly with lower anxiety levels in the mice.

The scientists then traced the exact physical pathway these signals travel. The electrical impulses move from the olfactory sensory neurons in the nose to mitral cells in the olfactory bulb, which acts as the brain’s smell-processing center. From there, the signals travel to a specific type of cell in the perirhinal cortex called parvalbumin-positive interneurons. These cells act like pacemakers, helping to synchronize brain waves. Finally, the pathway connects to the basolateral amygdala, an almond-shaped cluster of neurons known to play a massive role in processing fear and anxiety.

The researchers were struck by the precision of this mechanism. “First, how robust the effect is: nasal input frequency alone, delivered with no other cues, reliably shifts anxiety-like behavior in both directions,” Wu noted. “Second, the bidirectionality: the same circuit interprets slow input as calming and fast input as aggravating, working like a frequency dial for emotion. Third, the identity of a key relay: long-range projection interneurons in the perirhinal cortex; these PV+ cells act as a critical node in a limbic circuit.”

To verify that this exact pathway regulates anxiety, the researchers used chemogenetics. This method involves inserting engineered proteins into specific brain cells. These proteins do nothing until the animal is given a specialized designer drug, allowing researchers to temporarily turn specific brain pathways on or off without affecting the rest of the body.

When the team used the chemical drug to activate the parvalbumin-positive cells in the perirhinal cortex, the mice became less anxious and their high-gamma brain waves increased. Turning these cells off made the mice more anxious. This provided strong evidence that this specific chain of connections translates the speed of nasal signals into emotional states.

In a final phase of the research, the team tested whether slow breathing could act as a sustained treatment. They induced a chronic anxiety-like state in a group of mice using mild foot shocks. These anxious mice showed depleted high-gamma brain waves and avoided open spaces entirely.

The anxious mice then received a daily, two-hour treatment of slow stimulation for two weeks. The team tested both the light-based method and actual mechanical puffs of air delivered into the nasal cavity. Both the light and the air treatments completely restored the mice’s behavior to normal, pre-anxiety levels. The treatments also repaired the deficit in their high-gamma brain waves, and the calming effects lasted for at least a week after the daily treatments stopped.

The findings are in line with research covered by PsyPost in 2025, which found that an anxiety-inducing environment prompted rats to breathe faster, driving direct changes in brain wave frequencies. It is worth keeping in mind that the earlier study tracked natural, spontaneous breathing patterns in rats, while the current study actively manipulated nasal airflow to alter the emotional state of mice.

“The behavioral effects in mice were robust and reproducible, and a brief daily regimen of low-frequency nasal airflow produced lasting reductions in anxiety-like behaviors,” Wu said. “But this is a mouse study. The main value is establishing the principle and the mechanism: frequency-tuned nasal input is a legitimate handle on the anxiety system. Translation to humans will require clinical testing.”

As with all research, there are a few things to keep in mind. The study relied exclusively on male mice. Future research will need to include female mice to see if hormonal fluctuations change how this specific brain circuit operates.

Also, while the light-based stimulation provided an exact way to test the brain, it does not perfectly copy the complex physical sensation of air naturally flowing through a living animal’s nose. Because humans and mice have different brain structures, more research is needed to see if specific mechanical breathing devices or guided breathing exercises could target this same pathway to treat human anxiety disorders.

“This is a mouse study, so human benefits remain to be demonstrated,” Wu cautioned. “The effect depends on nasal airflow specifically, so mouth breathing would not engage this circuit; the advice should really be ‘breathe slowly through your nose.’ And this is a promising research direction in clinical anxiety.”

Looking ahead, the researchers have several next steps planned. “We are working in three directions: testing whether the same principles hold in humans, exploring whether the nose-brain axis is disrupted in disorders involving olfactory abnormalities, and studying how nasal stimulation shapes the brain, toward a practical, protocol-based nasal stimulation modulation,” Wu explained.

“One broader point: the nose may be an underappreciated gateway into the brain,” Wu concluded. “Because nasal input reaches limbic circuits, bypassing the thalamus, the nasal route offers a noninvasive and accessible window for regulating brain states. We hope this work encourages more attention to this direction.”

The study, “A nose-to-brain circuit underlies anxiety regulation by nasal afferent frequency in mice,” was authored by Xinsong Guo, Mengyan Liu, Qingcheng Xiong, Howai Ngai, Mingdong He, Xinying Li, Yingwei Zheng, Fuqiang Xu, Minghong Ma, and Ruiqi Wu.