Breath-holding is usually associated with swimming, freediving, or simple physical challenges. But researchers are increasingly interested in another possibility: whether the amount of time a person can voluntarily hold their breath may reveal something about how they respond to uncomfortable bodily sensations.
A new systematic review and meta-analysis published in Biological Psychology examined that question across 22 studies involving 1,263 adults. Researchers Maragda Puigcerver and Miguel Ángel Serrano of the Universitat de València investigated whether maximal voluntary breath-hold time could serve as a behavioral measure of interoceptive tolerance, particularly among people with anxiety disorders.
The findings are more specific than the idea that “people with anxiety cannot hold their breath for as long.” Adults with anxiety disorders generally demonstrated shorter voluntary breath-hold times than healthy controls, but the strongest evidence came from participants with panic disorder. The relationship was also much clearer for state anxiety and momentary distress than for stable trait anxiety.
That distinction matters because breath-holding is not simply a test of lung capacity. As the breath is held, carbon dioxide rises and the body generates increasingly noticeable sensations of air hunger and the urge to breathe. Researchers can therefore use the task as a controlled way to study how people tolerate and interpret uncomfortable internal signals.
Why Breath-Holding Is Interesting to Anxiety Researchers
Anxiety involves more than worrying thoughts. It can produce intense physical sensations, including changes in breathing, heart rate, muscle tension, dizziness, chest sensations, and a heightened awareness of internal bodily signals.
This is where interoception becomes relevant.
Interoception refers broadly to the brain’s perception and interpretation of signals originating inside the body. Researchers have studied it extensively in relation to anxiety because people experiencing anxiety or panic may interpret ordinary physiological sensations as threatening.
The new review focuses on a particularly measurable form of this process: how long someone voluntarily tolerates the increasingly uncomfortable sensations associated with not breathing.
The researchers describe maximal voluntary breath-hold time, or mvBHT, as a potential performance-based marker of interoceptive tolerance. Rather than asking someone to rate how anxious they feel, researchers can observe behavior during a controlled physiological challenge.
This approach complements conventional questionnaires. The National Institute of Mental Health anxiety information describes anxiety disorders as conditions involving excessive fear or anxiety that can affect behavior and physical functioning. A physiological challenge such as voluntary breath-holding gives researchers another way to investigate how those experiences interact with bodily signals.
The key question is therefore not whether someone can hold their breath for an impressive amount of time. It is whether their tolerance of respiratory discomfort is related to anxiety-related vulnerability.
What the New Meta-Analysis Included
Puigcerver and Serrano conducted a PRISMA-guided systematic review and meta-analysis, bringing together evidence from studies that used voluntary, unassisted breath-holding and measured anxiety-related psychological variables. The review was published online ahead of print on June 9, 2026, in Biological Psychology, volume 209.
The researchers included 22 studies with 1,263 adults. The studies examined both clinical anxiety groups and healthy comparison participants, with particular attention to panic disorder.
| Research Feature | Finding |
|---|---|
| Studies included | 22 |
| Total participants | 1,263 adults |
| Primary measure | Maximal voluntary breath-hold time |
| Main clinical focus | Anxiety disorders, especially panic disorder |
| Strongest association | State anxiety and momentary distress |
| Trait anxiety association | Marginal and not statistically significant after appropriate modeling |
The researchers also examined methodological differences between studies. One important factor was whether participants began the breath-hold after inhaling or after exhaling. Breath-holds initiated after inhalation tended to last longer in healthy samples than those initiated after exhalation.
That finding illustrates why breath-hold times cannot simply be compared between people without considering how the test was performed.
State Anxiety Was More Closely Linked to Breath-Hold Time
One of the most interesting results emerged when researchers separated trait anxiety from state anxiety.
Trait anxiety refers broadly to a relatively stable tendency to experience anxiety. State anxiety describes anxiety or distress occurring in a particular situation.
The review found only marginal, non-significant associations between maximal voluntary breath-hold time and trait anxiety once the statistical model accounted for dependence among effects within studies. The relationship became considerably clearer when researchers looked at state anxiety and momentary distress measured during or immediately after respiratory challenges.
For state anxiety and momentary distress, the pooled correlation was r = −0.19, with a 95% confidence interval from −0.29 to −0.09 and low heterogeneity across the included effects.
In practical terms, the negative relationship means that shorter breath-hold performance tended to be associated with greater momentary anxiety or distress.
But the size of the association matters.
A correlation of −0.19 is not evidence that breath-hold time can diagnose anxiety. It is a relatively small relationship that becomes meaningful mainly when considered alongside the wider pattern of findings.
Breath-Holding Is Really a Test of Interoceptive Tolerance
The central concept behind the research is not lung strength.
It is tolerance of internal discomfort.
During a voluntary breath-hold, the body continues producing carbon dioxide. As ventilation stops, chemical and mechanical signals increasingly push the person toward breathing again. The sensation can become uncomfortable well before an individual reaches anything resembling their absolute physiological limit.
That makes the task useful to researchers because participants have control over when to stop.
The decision to resume breathing can therefore reflect an interaction between physiological signals, perceived threat, attention, discomfort, and tolerance.
The review describes breath-holding as a way to generate a respiratory interoceptive challenge within controlled experimental conditions.
This is consistent with broader research examining how anxiety relates to the interpretation of bodily sensations. The American Psychological Association’s anxiety resources provide broader context for understanding anxiety as involving psychological and physiological responses rather than thoughts alone.
The new findings suggest that researchers may be able to use respiratory challenges to study this interaction more directly.
Why Breath-Hold Methodology Changes the Results
The review also highlights a problem that could easily be missed in headlines: there is no single universal breath-hold test.
Some studies begin the breath-hold after inhalation. Others begin after exhalation. Some use one maximal attempt, while others use repeated trials or standardized durations.
Those differences can materially affect performance.
The meta-analysis found that breath-hold onset was an important study-level predictor. In healthy samples, protocols beginning after inhalation produced longer breath-hold times than protocols beginning after exhalation.
This means a reported breath-hold duration cannot be interpreted properly without knowing the protocol.
The same principle applies to other physiological measurements. A number has meaning only within the conditions under which it was collected.
The review also found that the proportion of female participants predicted shorter average breath-hold times in healthy samples, while sex composition did not predict breath-hold time in panic-disorder samples.
These findings reinforce the need for standardized protocols before breath-hold time could become a widely used research or clinical measure.
People With Panic Disorder Showed the Clearest Difference
The strongest clinical finding involved panic disorder.
Adults with anxiety disorders generally demonstrated shorter maximal voluntary breath-hold times than healthy controls. However, the review found that this overall difference was driven primarily by a medium-sized reduction among people with panic disorder. For non-panic anxiety disorders, the effect was smaller and did not reach statistical significance.
This is important because panic disorder has a particularly strong relationship with fear of bodily sensations.
During a breath-hold, carbon dioxide gradually accumulates and the physiological drive to breathe becomes increasingly difficult to ignore. That experience can resemble one of the sensations that people with panic disorder may find threatening: the feeling that breathing is becoming difficult or insufficient.
Previous research has examined respiratory challenges in panic disorder, including the possibility that people with panic symptoms have heightened sensitivity to sensations associated with suffocation or rising carbon dioxide. The new meta-analysis does not establish a single biological explanation, but it strengthens the case for studying respiratory interoception in panic-related conditions.
For readers interested in breathwork and nervous-system regulation, this distinction is particularly useful. Controlled breathing practices and voluntary breath-holding are not interchangeable activities, and the new research should not be interpreted as evidence that deliberately holding the breath is a treatment for anxiety.
Does a Shorter Breath-Hold Mean Someone Has Anxiety?
No.
That is perhaps the most important practical point.
The researchers do not propose that people should test themselves at home by holding their breath and using the result to determine whether they have an anxiety disorder.
The evidence concerns group-level patterns across controlled research studies. It does not establish a diagnostic threshold.
Someone may stop a breath-hold early for many reasons, including unfamiliarity with the procedure, discomfort, respiratory fitness, expectations, experimental instructions, or simply choosing not to tolerate an unpleasant sensation.
The meta-analysis itself found that breath-hold time was much more consistently related to momentary distress than to stable trait anxiety.
That makes the measure potentially useful for researchers studying acute physiological discomfort, but much less suitable as a simple standalone indicator of someone’s long-term anxiety level.
It is also worth emphasizing that deliberate breath-holding can produce uncomfortable physiological changes. People with medical conditions affecting breathing, cardiovascular function, or consciousness should not use breath-holding exercises as an unsupervised experiment.
Why Panic Disorder Deserves More Research Attention
The stronger finding in panic disorder may help explain why respiratory sensations have received so much attention in panic research.
Panic attacks can involve sensations such as shortness of breath, choking sensations, chest discomfort, dizziness, and rapid heartbeat. For some individuals, those sensations can become part of a feedback loop in which a physical sensation is interpreted as evidence of danger, increasing fear and producing additional physiological arousal.
The breath-hold paradigm provides researchers with a way to introduce one type of respiratory discomfort under controlled circumstances.
The new meta-analysis found that the difference in breath-hold performance between anxiety-disorder groups and healthy controls was driven primarily by panic disorder rather than anxiety disorders as a broad category.
That specificity is scientifically valuable.
It suggests that future studies may need to distinguish between anxiety disorders rather than treating anxiety as one uniform condition.
What This Research Does Not Tell Us About Breathwork
The findings should not be confused with evidence that breath-holding itself improves anxiety.
The study examined breath-hold time as a measurement tool, not as a therapeutic intervention.
That is a major difference.
Breathwork programs typically involve controlled changes in breathing patterns, such as slower breathing, extended exhalation, or other structured techniques. Voluntary breath-holding, meanwhile, deliberately interrupts ventilation and creates increasing respiratory discomfort.
The physiological mechanisms and psychological experiences are therefore different.
Another 2026 systematic review examined the physiological effects of voluntary breath-holding and found that breath-hold conditions can significantly increase blood lactate, with substantial variation between protocols, training status, and apnea types. That review included 14 studies and 315 participants.
Together, these findings reinforce a broader point: respiratory practices should be studied according to their specific physiological demands rather than grouped together under the single label of “breathwork.”
The Bigger Shift Toward Measuring Interoception
The most interesting development may be the broader shift toward measuring how people experience their own bodies.
For years, much anxiety research depended heavily on questionnaires and interviews. Those tools remain important, but researchers increasingly want behavioral and physiological measurements that can complement self-reported symptoms.
Breath-hold time offers one possible measurement because it creates a controlled internal challenge.
A person receives increasingly strong respiratory signals, experiences discomfort, and chooses when to stop. Researchers can then compare that behavior with measures of anxiety, distress, fear of bodily sensations, and other psychological variables.
The new review suggests that this approach has some value, particularly for panic disorder and momentary anxiety-related distress.
It also shows why the field needs better standardization.
Different starting conditions can change performance. Different anxiety measures capture different aspects of the experience. And a measure that distinguishes people with panic disorder may not work equally well for generalized anxiety or other conditions.
That complexity makes the research more interesting, not less.
Where the Research Could Go Next
The next stage will likely involve determining whether breath-hold performance can do more than distinguish groups.
Researchers could investigate whether it changes alongside successful anxiety treatment, whether repeated measurements are stable within individuals, and whether respiratory interoceptive tolerance predicts future panic symptoms.
Another important question is whether interoceptive tolerance can be modified.
If researchers can demonstrate that a person’s response to respiratory sensations changes through evidence-based psychological or behavioral interventions, breath-hold paradigms could potentially become useful for studying treatment mechanisms.
For now, the evidence remains observational and experimental rather than diagnostic.
The 2026 review by Maragda Puigcerver and Miguel Ángel Serrano provides the strongest quantitative synthesis so far focused specifically on maximal voluntary breath-hold time and anxiety-related vulnerability. Its 22-study, 1,263-adult dataset suggests that shorter breath-hold performance is most clearly associated with panic disorder and with acute, state-level anxiety and distress.
That does not mean a person’s breath-hold time can tell them whether they are anxious.
It does suggest something more scientifically useful: the way people respond to uncomfortable respiratory sensations may provide researchers with another window into the relationship between the brain, body, and anxiety.
As interoception becomes a larger part of anxiety research, breathing may increasingly be studied not simply as something people can consciously control, but as a source of continuous information that the brain must interpret.
A New Way to Think About the Breath-Anxiety Connection
The emerging research changes the question.
Instead of asking whether people with anxiety simply “breathe differently,” researchers are examining how people experience, interpret, and tolerate the sensations produced by breathing.
The new systematic review provides evidence that voluntary breath-hold time can distinguish adults with anxiety disorders from healthy controls, with the strongest group difference appearing in panic disorder. It also shows that breath-hold performance is more consistently related to state anxiety and momentary distress than to stable trait anxiety.
That makes breath-holding an intriguing research tool, but not a home diagnostic test or a proven anxiety treatment.
The real significance lies in what happens when respiratory physiology and psychology are studied together. A breath-hold creates a measurable bodily challenge; the person’s response provides information about discomfort, attention, threat perception, and tolerance.
Future research will need to determine how reliable that information is and whether it can contribute to better assessment or treatment of anxiety-related conditions.
For now, the evidence points to a narrower but valuable finding: breath-hold performance appears to capture part of the relationship between respiratory sensations, acute distress, and anxiety vulnerability—especially in panic disorder.