AI Sheds New Light on IBS, Challenging One-Size-Fits-All Diagnosis

By using AI-driven cluster analysis, the research team identified at least eight distinct groups within those previously labelled as having IBS, revealing strong links between gastrointestinal symptoms and mental health.

Researchers in New Zealand have used artificial intelligence to re-examine irritable bowel syndrome and have uncovered evidence that patients currently grouped under the same diagnosis may, in fact, have very different underlying conditions.

The findings suggest that AI could help move IBS beyond its long-standing status as a “diagnosis by exclusion” and reduce years of uncertainty for patients.

Irritable bowel syndrome affects a significant proportion of the population and is among the most common gastrointestinal conditions worldwide. Despite often severe symptoms such as abdominal pain, bloating, diarrhoea, constipation and nausea, standard investigations including blood tests, stool analysis and colonoscopy frequently show no abnormalities.

The new study, led by the University of Auckland’s Bioengineering Institute and published in the journal Gut Microbes, applied machine learning techniques to an existing dataset of patients who underwent colonoscopy in Christchurch between 2016 and 2019.

Around 40 per cent of the 315 participants had been diagnosed with IBS, while a similar proportion had no gastrointestinal symptoms. By using AI-driven cluster analysis, the research team identified at least eight distinct groups within those previously labelled as having IBS.

“The crux of what we’ve found is that we have these big groups of patients we treat as having the same condition, but they don’t,” says lead researcher Dr Jarrah Dowrick. “Imagine your car doesn’t start and the diagnosis is ‘you have a bad car’. It’s overly reductive. It could be a dead battery, a bad starter motor, fuel problems, an electrical fault, or any number of other causes. In the same way, irritable bowel syndrome likely encompasses multiple different conditions.”

This oversimplification has real consequences. After receiving an IBS diagnosis, patients are often guided through prolonged trial-and-error treatment pathways, including dietary modification, psychological therapies and medications aimed at symptom control. When treatments fail, some patients report being told their symptoms are psychosomatic.

The AI analysis revealed that some patient clusters show strong links between gastrointestinal symptoms and mental health, while others appear largely gut-focused. “We’ve identified groups that are predominantly brain-centric and some that are predominantly gut-centric,” says co-author Dr Tim Angeli-Gordon. “Traditionally, patients with IBS would often be treated the same.” This distinction helps explain why stress-focused therapies may help some patients but offer little benefit to others.

Beyond IBS, the study underscores the broader promise and limits of AI in digital health. “The foundation of this paper is finding the most appropriate AI tool – in this case, cluster analysis – and then applying it to answer medical questions,” Dowrick says. He also cautions that AI depends on large, high-quality datasets, which require long-term investment, collaboration and patient participation.

“As machine learning becomes more powerful, we will see it used to gain deeper insights into more and more medical mysteries – but only if we also improve our ability to generate high-quality data,” he says.

The findings align with growing international debate around IBS classification. A 2025 paper from the Mayo Clinic suggested that the lack of clear biomarkers may eventually force clinicians to reconsider whether IBS should remain a single diagnosis. The New Zealand research adds further depth by showing how AI can detect clinically meaningful subtypes that were previously invisible.

Stay tuned for more such updates on Digital Health News