Artificial intelligence-based personalized diet: A pilot clinical study for irritable bowel syndrome

Area of Science:

• Gastroenterology
• Microbiome research
• Artificial Intelligence in Medicine

Background:

• Irritable Bowel Syndrome with Mixed features (IBS-M) is a common gastrointestinal disorder.
• Current management often involves dietary interventions, but personalized approaches are needed.
• Gut microbiome alterations are implicated in IBS pathogenesis.

Purpose of the Study:

• To evaluate the efficacy of an AI-based personalized nutrition diet in improving IBS-M symptoms.
• To assess the impact of personalized nutrition on the gut microbiome composition.
• To compare the outcomes of a personalized nutrition diet with a standard IBS diet.

Main Methods:

• Enrolled 25 IBS-M patients (Rome IV criteria).
• Collected fecal samples pre- and post-intervention for 16S rRNA sequencing.
• Assigned 14 patients to a 6-week personalized nutrition diet (AI-driven) and 11 to a standard IBS diet.
• Evaluated symptom severity using the Irritable Bowel Syndrome Severity Scoring System (IBS-SSS).

Main Results:

• Both groups showed significant IBS-SSS improvement (p < .02 for standard, p < .001 for personalized).
• 78% of the personalized nutrition group shifted from severe to moderate IBS-SSS scores, versus none in the standard diet group.
• Personalized nutrition increased beneficial genera: Faecalibacterium (p = .04), Bacteroides, and Propionibacterium.

Conclusions:

• AI-based personalized nutrition significantly improves IBS-M symptoms.
• Personalized diets modulate the gut microbiome, enhancing beneficial bacteria.
• Further large-scale, randomized, placebo-controlled trials are warranted to confirm long-term efficacy.