AI Identifies Distinct IBS Subtypes: University of Auckland Research Published in Gut Microbes

The Growing Burden of Irritable Bowel Syndrome in New Zealand

Irritable Bowel Syndrome (IBS), a common disorder of gut-brain interaction characterized by recurrent abdominal pain associated with altered bowel habits such as diarrhea, constipation, or both, affects approximately one in seven New Zealanders. Women are more than twice as likely to be diagnosed as men, making it a significant public health concern in the country. 71 70 This prevalence underscores the urgency for better diagnostic tools and treatments, as current approaches often rely on a 'diagnosis of exclusion' after ruling out other conditions through invasive tests like colonoscopies. Patients endure years of trial-and-error therapies, including dietary modifications, antispasmodics, laxatives, and psychological interventions, with limited success for many.

In New Zealand, where healthcare resources are stretched, IBS contributes to substantial economic costs through lost productivity and medical visits. The condition's heterogeneity—varying symptoms, triggers, and responses—has long puzzled clinicians. Traditional classifications divide IBS into subtypes like IBS with constipation (IBS-C), diarrhea (IBS-D), mixed (IBS-M), and unsubtyped (IBS-U), based on predominant bowel habits. However, these categories fail to capture underlying biological differences, leading to suboptimal outcomes.

University of Auckland Leads Breakthrough with AI-Powered Analysis

The University of Auckland, New Zealand's leading research institution, has pioneered a transformative approach through its Auckland Bioengineering Institute. Researchers Dr. Jarrah Dowrick and Dr. Tim Angeli-Gordon spearheaded a study published in the prestigious journal Gut Microbes, leveraging artificial intelligence (AI) to dissect IBS complexity. 71 This work exemplifies how New Zealand universities are at the forefront of integrating engineering, biology, and medicine to address real-world health challenges.

The study, titled "Integrated multi-omic and symptom clustering reveals lower-gastrointestinal disorders of gut-brain interaction heterogeneity," analyzed data from the COMFORT cohort—a Christchurch-based study spanning 2016 to 2019. By applying unsupervised machine learning, the team uncovered patterns invisible to conventional methods, positioning the University of Auckland as a hub for innovative research jobs in bioengineering and gastroenterology.

Building the Foundation: The COMFORT Cohort Dataset

The foundation of this research was a robust dataset from 315 participants undergoing colonoscopy in Christchurch. Approximately 40% had IBS diagnoses (around 126 individuals), 40% were symptom-free controls, and the rest had other gastrointestinal issues. Participants provided comprehensive data: symptom questionnaires (Rome IV criteria, Hospital Anxiety and Depression Scale or HADS, Symptom Abnormality in Gastro-Intestinal Disorders Score or SAGIS, Patient-Reported Outcomes Measurement Information System or PROMIS, and Food and Symptom Time-diary or FAST), shotgun metagenomics for gut microbiome profiling, untargeted metabolomics from plasma and feces, and targeted assays for amino acids and bile acids/short-chain fatty acids. 72

This multi-omic approach—combining symptoms, genetics, microbiome, and metabolites—enabled a holistic view. Collecting and curating this data took nearly a decade, involving collaborators across New Zealand universities like Otago and Massey, highlighting the collaborative spirit in Kiwi higher education.

How AI Cluster Analysis Revolutionized IBS Subtyping

At the core was unsupervised machine learning via cluster analysis. First, symptoms were factor-analyzed into consolidated scores (e.g., pain, bloating, depression, anxiety, diarrhea, constipation). Latent profile analysis grouped patients by symptoms, while the NEMO pipeline clustered biological data. These were integrated using consensus matrices and hierarchical clustering, with stability validated through 1,000 bootstraps (clusters stable if pair-wise co-clustering >50%). The Davies-Bouldin Index and silhouette scores optimized cluster numbers. 72

This step-by-step process revealed 11 clusters, eight stable ones. Unlike Rome IV's symptom-based subtypes, AI integrated biology, exposing true heterogeneity. For aspiring researchers, such methodologies open doors to academic careers in AI-health intersections.

Photo by Adrien Olichon on Unsplash

• Factor analysis reduces multidimensional symptoms to key factors.
• Biological clustering identifies microbiome/metabolite patterns.
• Integrated merging uncovers multimodal subtypes.
• Bootstrapping ensures reproducibility.

Discovering Eight Distinct IBS Subtypes

The AI analysis identified at least eight distinct clusters within IBS-like disorders of gut-brain interaction (DGBI). Notable subtypes include:

• Gut-Centric, High Pain (Cluster 1): High constipation, pain, bloating, some diarrhea; elevated microbiome diversity but low short-chain fatty acids (SCFAs) and bile acids.
• Brain-Centric (Clusters 7,8): High anxiety/depression linked to symptoms; minimal gut biology alterations.
• Dysbiotic DGBI (Cluster 11): Low symptoms but high Firmicutes/Bacteroidetes ratio, low richness, elevated branched-chain amino acids; suggests barrier dysfunction.

Read the full study in Gut Microbes for detailed cluster profiles. 72 These findings challenge the one-size-fits-all model, as no cluster aligned strictly with IBS-C/D/M. 71

Gut Microbiome's Pivotal Role in IBS Heterogeneity

The microbiome emerged as key differentiator. Gut-Centric clusters showed disrupted ammonia handling: upregulated amino acid degradation (carbon/lysine) but downregulated urea cycle, potentially lowering glutamate/GABA for pain modulation. Dysbiotic types had enriched transport systems (ABC transporters) and opportunistic bacteria, possibly from high-sugar diets impairing gut barriers. SCFAs, vital for gut health, were depleted in high-pain groups. 72

This aligns with global evidence of dysbiosis in post-infectious IBS, emphasizing New Zealand's strength in microbiome research at institutions like Auckland.

Transforming Treatment: From Trial-and-Error to Precision Medicine

Subtype insights promise targeted therapies. Brain-centric patients may benefit from cognitive behavioral therapy or antidepressants, while gut-centric ones need microbiome modulators like pre/probiotics or low-FODMAP diets. Dysbiotic groups could target barrier repair or nitrogen metabolism.PMC full text Clinicians in New Zealand can now stratify patients, reducing frustration quoted by Dr. Dowrick: "Imagine your car doesn’t start and the diagnosis is ‘you have a bad car’—overly reductive."

• Brain-centric: Psychological interventions first.
• Gut-centric: Diet/microbiome therapies.
• Dysbiotic: Antimicrobial or metabolite-targeted drugs.
• Benefits: Faster relief, lower costs, better quality of life.

Implications for New Zealand's Healthcare and Higher Education

For Kiwi patients, this means less misattribution to 'stress alone.' Universities like Auckland drive this via bioengineering programs, fostering interdisciplinary talent. The study's decade-long effort highlights data-sharing needs across NZ institutions, potentially informing national guidelines.

Explore university opportunities in New Zealand or higher ed jobs to contribute to such research.

Photo by Bill Fairs on Unsplash

Future Outlook: AI's Expanding Role in Gastroenterology

Looking ahead, larger datasets from routine clinical integration will refine subtypes, identifying biomarkers for non-invasive tests. University of Auckland calls for data investments to unlock AI's potential in other mysteries. Related NZ work, like Otago's GUTS unit, complements this. For students, this signals booming demand for research assistant jobs in AI-gut health.

Stakeholders—patients, doctors, policymakers—gain actionable insights: advocate for multi-omic screening, fund microbiome trials, pursue personalized care.

Career Pathways in IBS and Microbiome Research

This breakthrough opens doors in New Zealand higher education. From PhDs in bioengineering to faculty positions, opportunities abound. Check university jobs, rate professors, or career advice on AcademicJobs.com. With global IBS burden, NZ researchers lead the charge.