Every individual possesses a distinct pattern of nasal breathing, which remains constant over extended periods and acts as a biological fingerprint. A recent study, which monitored people’s inhalations and exhalations across a full twenty-four hours, shows that these unique respiratory signatures enable the identification of individuals with near-perfect accuracy, in addition to forecasting personal levels of anxiety, depression, and body mass index. The findings of this research were published in the journal Current Biology.
Breathing frequently appears to be an automatic and simple physiological action. Most people only become aware of their respiration when struggling for air or engaging in strenuous physical activity. Nevertheless, the entire process of drawing air in and pushing it out is governed by an extensive and complex neural network.
This neural system primarily operates within the brainstem. It functions much like a biological pacemaker, continuously modulating a person’s breathing to match current physiological demands. The system gathers vast amounts of sensory input from across the body, adjusting the rate and depth of every single breath.
Given the inherent uniqueness of the human brain in both its structure and electrical activity, the researchers hypothesized that the biological outputs generated by these localized brain circuits might also exhibit a high degree of individuality. To test this concept, scientists at the Weizmann Institute of Science in Israel devised an experiment designed to track the movement of air through the nose with great precision over a long duration.
The lead investigators, Timna Soroka and Noam Sobel, opted to focus their study on the nose rather than the mouth. Nasal passages maintain a special link with the brain because they are densely packed with sensory nerves that constantly report on air flow. The brain actively manages this process, systematically alternating which nostril takes the lead in the breathing cycle.
To capture these long-term respiratory patterns, the team engineered a specialized wearable device. This compact tracker was affixed to the back of the volunteer’s head and connected to a nasal cannula—a thin plastic tube featuring two small prongs that were positioned directly inside the nostrils.
Unlike typical clinical examinations which assess breathing for only a few minutes to gauge lung capacity, this setup continuously recorded respiration throughout both day and night. The device incorporated highly sensitive pressure sensors, which independently measured airflow for the left and right nostrils in real time. It logged data at a rate of six times per second, capturing the most minute dynamic variations in air movement.
Approximately one hundred healthy participants, mostly between the ages of twenty and thirty, took part in the research. Each subject wore the tracker for a complete twenty-four-hour period, going about their normal routines and logging their major activities and sleep patterns using a provided smartphone app.
For a subset of over forty participants, the research team replicated the entire procedure. These individuals returned to wear the recording apparatus for a second twenty-four-hour session. The interval between the first and second recording phases ranged from a few days to nearly two years.
When the scientists input the raw respiratory data into a computational model, they discovered they could identify individual subjects with remarkable success. Based solely on awake-state breathing data, the system correctly pinpointed specific people within the group with 96.8% accuracy.
The success rate of this identification process places respiratory patterns on par with established biometric measures like voice recognition. The findings indicated that human breathing is not merely a generalized mammalian rhythm but rather a personalized behavioral characteristic.
This capacity to recognize a person through their breathing style remained robust even across significant time gaps. Even when a computer model learned an individual’s breathing pattern during the initial testing day, it could successfully pick them out from a crowd using data collected up to twenty-three months later.
To confirm that the computer was analyzing the actual act of breathing, rather than just general patterns of physical movement, the researchers also examined data from a motion sensor integrated into the device. While analyzing body movements permitted some level of breath identification, its accuracy was markedly lower than that achieved through nasal airflow analysis.
To validate these results, the investigators assessed dozens of distinct metrics within the breathing data. They organized the information into measures such as inhaled volume, the duration of pauses between breaths, and the asymmetry of airflow between the left and right nostrils.
No single feature in isolation could distinguish one person from another. High overall identification accuracy required the computational model to simultaneously analyze approximately twenty to one hundred different respiratory characteristics.
Beyond simple personal identification, the researchers explored whether these respiratory traits revealed anything about a person’s physical condition. They analyzed the baseline data to look for physiological markers, such as the transition between waking and sleeping states.
The analysis showed clear shifts between being awake and being asleep. As participants drifted off, the total volume of air inhaled and exhaled decreased, while the switching dominance between the right and left nostrils became more pronounced. By examining just five minutes of a person’s breathing data, the model could readily determine if they were asleep or awake.
The continuous airflow data also mathematically correlated with the participants’ Body Mass Index (BMI), a standard calculation based on a person’s height and weight. The research team observed a mathematical link between an individual’s body mass and specific aspects of their nasal cycle, suggesting that the neural dynamics governing respiration directly interact with body composition.
In addition to tracking physical metrics, the researchers sought to ascertain if these breathing profiles reflected certain aspects of human cognition and emotion. All participants completed standard psychological questionnaires to gauge their baseline levels of anxiety, reported depressive symptoms, and behavioral traits associated with autism.
Although the group studied consisted of otherwise typical adults without major clinical diagnoses, the measured breathing patterns showed correlations with the survey outcomes. The researchers found they could partially predict a person’s score on a depression scale based exclusively on their breathing specifics, such as the peak inhalation speed during waking hours.
Similar predictive relationships were noted for general anxiety. Participants who scored higher on trait anxiety scales tended to exhibit somewhat shorter inhalation durations while asleep. Minor variations in the length of pauses between breaths were also linked to varying self-reported levels of anxiety.
Analysis of the autism spectrum questionnaire data again revealed mathematical associations with the participants’ breathing. Subtle changes in the duration of pauses during inhalation corresponded to different measures of behavior. These results imply that emotional and cognitive states leave faint but discernible biological traces in how the brainstem regulates respiration.
While the study offers a novel way to measure fundamental human biological parameters, the experimental method has a few notable drawbacks. The nasal cannula occasionally shifted during the participants’ sleep, which caused interruptions in nighttime data collection.
Furthermore, the pressure sensors placed inside the nose are excellent at recording the precise timing of breaths but lack absolute precision when calculating the total volume of air reaching the lungs. The appearance of the apparatus might also limit its practical application for everyday use, as wearing medical tubing across the face is highly visible.
Looking ahead, the researchers intend to expand the use of this testing methodology to broader populations. Since breathing characteristics offer direct insight into brain function, the team anticipates applying this approach to investigate various diseases. Monitoring a patient’s unique respiratory profile over time could ultimately serve as a passive, non-invasive tool for tracking overall neurological wellness.
About the Author
