From Step Counters to Gut Health Co-Pilots

Many of us wear tiny research labs on our wrists, fingers, and even shoes. These devices quietly track our steps, heartbeats, and sleep cycles. But the future of wearables isn’t just counting steps, it’s building personalized health co-pilots: adaptive, AI-powered systems that understand your unique rhythms and predict problems before they happen. For people with chronic gastrointestinal (GI) conditions like IBS, this shift could be transformative.

This announcement aligns with Health Secretary’s June 24 plan to launch a nationwide campaign encouraging every American to adopt wearables, part of his ambitious “Making America Healthy Again” agenda.

A recent review highlights how wearable devices, from smartwatches to sensor patches, are already being used to monitor IBD activity, predict flares, and track biomarkers like fecal calprotectin and C-reactive protein. Future applications may include ingestible sensors, microbiome monitoring, and machine learning-driven early disease detection. Similarly, advances in acoustic sensing are turning bowel sounds into a diagnostic tool for IBS. Miniaturized microphones, AI models like convolutional and recurrent neural networks, and portable recording devices are bringing continuous, objective GI monitoring closer to reality.

But there’s a deeper layer emerging beneath these devices: Network Medicine (NM). NM maps disease as disruptions in interconnected molecular and physiological networks rather than single gene or biomarker abnormalities. By treating diseases like IBS as network-wide phenomena, NM provides a framework for understanding how gut inflammation links to immune responses, microbiome shifts, and even stress hormones. When combined with AI, especially deep learning, NM allows researchers to integrate massive multi-omic datasets (genomics, proteomics, metabolomics) with wearable device streams, revealing subtle, individualized signatures of disease progression or recovery. This fusion moves beyond simple symptom tracking, creating biologically grounded, predictive health models.

—echoing earlier analyses of AI’s underutilization in biomedicine—underscore this theme: despite rapid advances, there are persistent gaps between benchmark performance and real-world usability, emphasizing the need for dynamic, evaluator-aware frameworks to guide safe clinical adoption.

Similarly, NM’s predictive networks must adapt dynamically to individual variability. Researchers are working on cross-user adaptive AI and network-aware modeling that runs efficiently on-device, preserving privacy while continuously refining predictions. Recent systematic reviews of large language models in healthcare - echoing earlier analyses of AI’s underutilization in biomedicine - underscore this theme: despite rapid advances, there are persistent gaps between benchmark performance and real-world usability, emphasizing the need for dynamic, evaluator-aware frameworks to guide safe clinical adoption.

Imagine a wearable ecosystem that does more than log symptoms - it connects the dots through network science: a smartwatch detecting subtle heart rate variability, a ring tracking skin temperature trends, and an ingestible sensor analyzing gut pH. Integrated with NM-driven AI, these signals could identify emerging disease network perturbations and predict a flare-up 48 hours in advance, guiding personalized diet tweaks, medication adjustments, or stress management before symptoms strike. Bowel sound analysis, combined with unobtrusive “toilet-lab” technology could replace tedious food diaries with objective, automated insights, creating a continuous and rich feedback loop between the body and the wearer.

The leap from step counters to gut health co-pilots isn’t just a technical upgrade; it’s a paradigm shift -from reactive care to proactive, precision health. By merging wearable technology with AI and NM, supported by seamless in-home lab testing, we’re approaching a future where chronic GI condition management is informed by both real-time physiological data and deep network-level understanding of disease. With careful design, strong privacy safeguards, and adaptive AI, wearable tech could evolve into indispensable tools that don’t just monitor illness but actively shape health outcomes.

REFERENCES

Nicholas GO, Faith LI, Jeric MC, KO O, Kelvin KF, Seung-Min PA, Christopher HT, Sunny HW. Acoustic sensing and analysis of bowel sounds in irritable bowel syndrome-recent engineering developments and clinical applications. Sensors and Actuators A: Physical. 2025 Jul 22:116910.

Harindranath S, Desai D. Wearable technology in inflammatory bowel disease: current state and future direction. Expert Review of Medical Devices. 2025 Feb 1;22(2):121-6.

Irene S. Gabashvili Evaluating General-Purpose LLMs for Patient-Facing Use: Dermatology-Centered Systematic Review and Meta-Analysis medRxiv 2025.08.11.25333149; doi: https://doi.org/10.1101/2025.08.11.25333149

Erturk E, Kamran F, Abbaspourazad S, Jewell S, Sharma H, Li Y, Williamson S, Foti NJ, Futoma J. Beyond Sensor Data: Foundation Models of Behavioral Data from Wearables Improve Health Predictions. arXiv preprint arXiv:2507.00191. 2025 Jun 30.  https://arxiv.org/abs/2507.00191

Cai Y, Guo B, Salim F, Hong Z. Towards Generalizable Human Activity Recognition: A Survey. arXiv preprint arXiv:2508.12213. 2025 Aug 17. https://arxiv.org/abs/2508.12213

Harvard dropouts to launch 'always on' AI smart glasses that listen and record every. TechCrunch Aug 20, 2025 conversation. https://techcrunch.com/2025/08/20/harvard-dropouts-to-launch-always-on-ai-smart-glasses-that-listen-and-record-every-conversation/

Predicting Meal Responses: A Storm of Variables

Predicting how our bodies respond to meals—particularly for individuals with conditions like Irritable Bowel Syndrome - is notoriously challenging. Traditional approaches have relied heavily on averages and generalized dietary guidelines, but each individual's unique genetic makeup, gut microbiome, and lifestyle make standardized predictions unreliable. Aurametrix 1.0 was pioneering software designed to personalize predictions based on individual characteristics, acknowledging that our biology makes us distinct. Yet, as recent studies highlight, even this level of personalization might fall short.

A recent exploratory analysis published in the American Journal of Clinical Nutrition underscores the complexity of individual responses to identical meals. Researchers found substantial variability in glucose responses when the same meal was consumed by the same individual on different days. Using continuous glucose monitors (CGMs), the study tracked glucose responses in participants consuming identical meals one week apart under tightly controlled conditions. Remarkably, about 80% of the variation in glucose responses was attributed to intraindividual factors—biological differences, environmental conditions, and even measurement error—rather than the food itself.

This variability wasn't significantly explained by common factors such as carbohydrate content, energy intake, or physical activity alone. The timing of snack intake, water consumption, recent stress levels, sleep quality, and minor changes in meal composition or eating sequence also influence bodily responses significantly. Similar complexities occur in predicting symptom flares in IBS, where digestive responses can vary dramatically based on recent dietary patterns, hydration, psychological stress, and sleep history.

Behavioral and psychological factors, like stress or anxiety, can profoundly impact digestive function and gut sensitivity, causing variability in how someone with IBS responds even to familiar meals. Likewise, hydration status and recent dietary history can modulate the gut's reaction, making precise predictions difficult. Interestingly, while intraindividual variation (iiV) presents a challenge in the general population, individuals with metabolic disease—particularly those with type 1 or type 2 diabetes—tend to exhibit more predictable glycemic responses to meals - particularly when food categories—rather than just macronutrients—are properly structured through more sophisticated algorithms like Hierarchical Information Criterion (as suggested in Aurametrix 1.0 and demonstrated by Shen et al, 2025)

Emerging research further underscores that a person's response to a meal isn't static but rather dynamic, influenced by a complex interplay of factors over days and weeks, not just hours. These findings suggest that true personalized nutrition must account not only for an individual’s static biological markers but also dynamic behavioral and environmental factors, including past dietary habits, hydration levels, stress, sleep patterns, and even subtle daily activity variations.

Thus, while personalized prediction software like Aurametrix's version 1.0 represented a significant leap forward, the next generation of personalized nutrition and IBS management tools must integrate broader and more comprehensive real-time data. Only then can we better anticipate and manage complex, highly individualized bodily responses.

REFERENCES

Hengist A, Ong JA, McNeel K, Guo J, Hall KD. Imprecision nutrition? Intraindividual variability of glucose responses to duplicate presented meals in adults without diabetes. Am J Clin Nutr. 2025 Jan;121(1):74-82. doi: 10.1016/j.ajcnut.2024.10.007. Epub 2024 Dec 2. PMID: 39755436; PMCID: PMC11747189.

Wolever TM. Personalized nutrition by prediction of glycemic responses: garbage in → garbage out. Am J Clin Nutr. 2025 Jan;121(1):1-2. doi: 10.1016/j.ajcnut.2024.11.004. Epub 2024 Dec 2. PMID: 39755431.

Shen Y, Choi E, Kleinberg S. Predicting Postprandial Glycemic Responses With Limited Data in Type 1 and Type 2 Diabetes. J Diabetes Sci Technol. 2025 Mar 5:19322968251321508. doi: 10.1177/19322968251321508. Epub ahead of print. PMID: 40042044; PMCID: PMC11883769.

A New Way to Track Dietary Intake?

Diet influences almost every aspect of our health—from digestion and mood to energy levels and chronic disease risk. Yet accurately tracking what we eat has always been tricky. Self-report methods, like food diaries and questionnaires, often fall short: Who remembers exactly what they ate two days ago? Was it one glass of orange juice or two?

But what if the answer to tracking your diet - and, by extension, better managing Irritable Bowel Syndrome (IBS) - was as simple as a smart toilet analyzing your stool?

In a new study published in Nature Metabolism, researchers from the Institute for Systems Biology introduce Metagenomic Estimation of Dietary Intake (MEDI) — a novel approach to dietary analysis that relies on the genetic material found directly in human stool. By analyzing food-derived DNA present in fecal samples, MEDI precisely identifies and quantifies dietary components without the inaccuracies of self-reporting. Rather than relying on memory or written logs, MEDI tracks dietary intake by detecting and measuring tiny fragments of food-derived DNA in stool samples.

The research team first created a vast database containing genetic fingerprints of over 400 edible plants and animals. They then developed a specialized algorithm capable of spotting traces of this DNA within the complex mixture of human and microbial genetic material found in stool samples.

Although food DNA makes up only about 0.001% of the total DNA in stool, MEDI accurately identifies these dietary traces. In controlled feeding studies, MEDI’s accuracy rivaled detailed daily food diaries, correctly estimating calorie, protein, carbohydrate, potassium, cholesterol, and vitamin B12 intake.

This precision can offer invaluable insights to IBS sufferers, who often struggle to pinpoint specific dietary triggers behind their symptoms.

MEDI accurately identified when babies transitioned from milk to solid food, beginning around 160 days of age. MEDI’s dietary estimates closely matched results from traditional food frequency questionnaires, validating its real-world effectiveness. MEDI even identified specific dietary patterns linked to metabolic syndrome—highlighting higher animal-food intake and lower plant-based consumption, along with increased lactose, cholesterol, and certain fats. These findings reinforce well-known dietary patterns linked to inflammation and chronic gastrointestinal discomfort, common in IBS.

MEDI’s potential to objectively track diet offers hope for IBS patients, potentially revealing personalized dietary interventions without tedious logging. As research evolves, IBS patients should keep a close eye on stool analysis approaches —it just might hold the key to better dietary management and improved quality of life.

Diet is the most important piece of the puzzle in IBS. Recent advances in treatments provide more drug-based solutions. In addition to currently available treatments, such as laxatives, antidiarrheals, analgesics, and antispasmodics, targeting the underlying stress with opioid delta-receptor agonists (DOP agonists) may offer a quicker solution with minimal adverse effects. Animal studies showed these compounds reduced abdominal pain and improved bowel regularity through their influence on the brain’s insular cortex, potentially relieving both physical and emotional symptoms. Other promising solutions are psychotherapy, including cognitive-behavioral therapy (CBT) and psychodynamic therapy, Ibodulant, targeting gut neurokinin receptors, significantly reducing abdominal pain and diarrhea symptoms in early studies; GaRP (Gastrointestinal Reprogramming) aiming to reset gut functions, and Blautix, a live probiotic therapy designed to restore gut microbiome balance.

REFERENCES

Diener C, Holscher HD, Filek K, Corbin KD, Moissl-Eichinger C, Gibbons SM. Metagenomic estimation of dietary intake from human stool.  Nat Metab. 2025 Feb 18. doi: 10.1038/s42255-025-01220-1. Online ahead of print. PMID: 39966520 (preprint: bioRxiv 2024 Feb 6:2024.02.02.578701. doi: 10.1101/2024.02.02.578701.

Yoshioka, T., et al. (2024). Agonists of the opioid δ-receptor improve irritable bowel syndrome-like symptoms via the central nervous system. British Journal of Pharmacology. doi.org/10.1111/bph.17428.

Mozaffari S, Nikfar S, Abdollahi M. Drugs of the future for diarrhea-predominant irritable bowel syndrome: an overview of current investigational drugs. Expert Opinion on Investigational Drugs. 2024 Mar 3;33(3):219-28.

Pitashny M, Kesten I, Shlon D, Hur DB, Bar-Yoseph H. The Future of Microbiome Therapeutics. Drugs. 2025 Jan 23:1-9.

Brian Doctrow, Tracking diet from stool samples. https://www.nih.gov/news-events/nih-research-matters/tracking-diet-stool-samples