Researchers at University College London placed a digital stethoscope on 45 people's abdomens, played them two 15-second recordings of their own gut noises, and asked which one was happening right now.
Participants got it right roughly 56% to 60% of the time. A coin flip gets 50%.
That result is not a failure. It is the point of a new test called Rumble Recognition, published in Scientific Reports, designed to measure something researchers have struggled to quantify: how accurately a person reads signals coming from their own digestive system.
Two Clips, One of Them Live
The design is deceptively simple. On each trial, a participant hears two audio clips streamed from a digital stethoscope on the abdomen. One is live. The other was recorded minutes earlier. They pick which is which, then rate their confidence on a sliding scale running from guess to certain.
Each person completed 10 trials in each of three physiological states: after a three-hour fast, after drinking sparkling water, and after consuming a high-protein meal replacement shake. The sequence was chosen deliberately, since carbonation and a protein load change how much a stomach has to say. The threshold for beating chance was set at eight or more correct out of 10, which corresponds to a binomial probability of roughly 0.05.
The sample was 45 healthy adults, 68% of them women. Ethical approval came from the UCL Human Research Ethics Committee, and the work was funded through a Wellcome Mental Health Award alongside Medical Research Council and NIHR support.
The Interesting Result Was Who Beat Chance
Average accuracy hovering just above 50% would look like a dead end if the team was trying to prove people can identify their own gut. They were not. They were looking for variation between individuals, and they found it.
Roughly one in five participants performed significantly above chance while fasting. That rose to about one in four after drinking either the sparkling water or the shake, which makes physiological sense: a stomach that has just received something has more to signal.
Some participants also showed genuine metacognitive insight, meaning their confidence ratings tracked whether they were actually right. That is a separate ability from raw accuracy, and it is the one clinicians tend to care about, since it reflects whether a person can tell when their own bodily read is trustworthy.
Why Build a Test Out of Gut Noise
Interoception, the sense of your own internal state, has become a substantial area of neuroscience because it underpins hunger, thirst, bladder fullness and, the researchers argue, emotional experience. Gastrointestinal sensations contribute to what people casually call gut feelings.
The problem is measurement. Established interoception tasks are overwhelmingly cardiac: count your heartbeats, judge whether a tone is synchronized with your pulse. Existing gut methods are far less pleasant, involving inflatable balloons or swallowed electronic capsules, which makes recruitment difficult. A less invasive probe of gut mechanosensation has been demonstrated using an ingestible vibrating pill, but it still requires swallowing hardware.
That gap matters clinically. Interoceptive differences have been documented across functional gastrointestinal disorders, and the authors argue their task could eventually help study eating disorders, irritable bowel syndrome, anxiety and depression. The task is fully automated, needs only modest equipment, and the code is freely available as open source.
The approach has a precedent. A 2016 study found that people could identify recordings of their own heartbeat above chance while showing poor insight into when they were correct, suggesting implicit knowledge of bodily signals that does not reach conscious awareness. Rumble Recognition asks whether the same holds for the gut.
A Proof of Concept, Not a Diagnostic
The limits are substantial, and the authors present the work as proof of principle rather than a finished instrument.
Forty-five healthy adults is a small, non-clinical sample. The study did not include people with eating disorders, IBS or inflammatory bowel disease, which is where the test would eventually need to prove itself. The paper was also posted as an unedited version to give early access to the findings, meaning further editing is still to come.
There is also a conceptual caveat worth stating. Performing poorly on this task does not mean a person is disconnected from their body or that anything is medically wrong. Gut sounds, known clinically as borborygmi, are generated by muscular contractions moving contents through the intestines, and hearing them from the outside is a genuinely unusual perceptual demand.
Anyone with persistent abdominal pain, changes in bowel habits, or unexplained digestive symptoms should see a clinician rather than try to interpret their own internal soundtrack. The National Institute of Diabetes and Digestive and Kidney Diseases maintains guidance on when digestive symptoms warrant evaluation.
Key Questions Answered
What is the test?
Rumble Recognition plays participants two 15-second recordings of their own gastrointestinal sounds through a digital stethoscope, one live and one recorded minutes earlier, and asks them to identify the live one.
How well did people do?
Accuracy averaged about 56% to 60%, only modestly above the 50% expected from guessing.
Did anyone perform well?
Yes. Roughly one in five beat chance while fasting, rising to about one in four after drinking sparkling water or a protein shake.
What is gastrointestinal interoception?
The ability to sense, interpret, and integrate signals coming from the digestive system, a component of what people describe as gut feelings.
Why does this matter for health?
Interoceptive differences have been documented in eating disorders and gut conditions, but researchers have lacked a standardized, non-invasive way to measure the gastrointestinal component specifically.
Should I worry if I could not do this?
No. The task is perceptually difficult by design; the sample was healthy volunteers, and performance has no established clinical meaning.
