Harnessing Large Language Models to Uncover Insights in Diabetes Wearable Data

Harnessing Large Language Models to Uncover Insights in Diabetes Wearable Data | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Harnessing Large Language Models to Uncover Insights in Diabetes Wearable Data Arash Alavi, Kexin Cha, Delara P Esfarjani, Bhavesh Patel, Jennifer Li Pook Than, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4966049/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Large Language Models (LLMs) have gained significant attention and are increasingly used by researchers. Concurrently, publicly accessible datasets containing individual-level health information are becoming more available. Some of these datasets, such as the recently released Artificial Intelligence Ready and Equitable Atlas for Diabetes Insights (AI-READI) dataset, include individual-level data from digital wearable technologies. The application of LLMs to gain insights about health from wearable sensor data specific to diabetes is underexplored. This study presents a comprehensive evaluation of multiple LLMs, including GPT-3.5, GPT-4, GPT-4o, Gemini, Gemini 1.5 Pro, and Claude 3 Sonnet, on various diabetes research tasks using diverse prompting methods to evaluate their performance and gain new insights into diabetes and glucose dysregulation. Notably, GPT-4o showed promising performance across tasks with a chain-of-thought prompt design (aggregate performance score of 95.5%). Moreover, using this model, we identified new insights from the dataset, such as the heightened sensitivity to stress among diabetic participants during glucose level fluctuations, which underscores the complex interplay between metabolic and psychological factors. These results demonstrate that LLMs can enhance the pace of discovery and also enable automated interpretation of data for users of wearable devices, including both the research team and the individual wearing the device. Meanwhile, we also emphasize the critical limitations, such as privacy and ethical risks and dataset biases, that must be resolved for real-world application in diabetes health settings. This study highlights the potential and challenges of integrating LLMs into diabetes research and, more broadly, wearables, paving the way for future healthcare advancements, particularly in disadvantaged communities. Biological sciences/Computational biology and bioinformatics/Machine learning Health sciences/Diseases/Endocrine system and metabolic diseases/Diabetes Figures Figure 1 Figure 2 Figure 3 Full Text Additional Declarations Competing interest reported. Dr. Snyder is a cofounder and scientific advisor of Personalis, SensOmics, Qbio, January AI, Fodsel, Filtricine, Protos, RTHM, Iollo, Marble Therapeutics, Crosshair Therapeutics, NextThought, and Mirvie. He is also a scientific advisor of Jupiter, Neuvivo, Swaza, Mitrix, Yuvan, TranscribeGlass, and Applied Cognition. Dr. Lee reports grants from Santen, personal fees from Genentech, personal fees from US FDA, personal fees from Johnson and Johnson, personal fees from Boehringer Ingelheim, non-financial support from iCareWorld, grants from Topcon, grants from Carl Zeiss Meditec, personal fees from Gyroscope, non-financial support from Optomed, non-financial support from Heidelberg, non-financial support from Microsoft, grants from Regeneron, grants from Amazon, grants from Meta, outside the submitted work. The other authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4966049","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":360679190,"identity":"70d6d452-cead-4059-961b-8ee128a8c03a","order_by":0,"name":"Arash Alavi","email":"","orcid":"","institution":"Stanford University","correspondingAuthor":false,"prefix":"","firstName":"Arash","middleName":"","lastName":"Alavi","suffix":""},{"id":360679191,"identity":"95ea79dc-bd69-438a-9005-62af97ad6fd5","order_by":1,"name":"Kexin Cha","email":"","orcid":"","institution":"Stanford University","correspondingAuthor":false,"prefix":"","firstName":"Kexin","middleName":"","lastName":"Cha","suffix":""},{"id":360679193,"identity":"5dbcd719-2131-45e0-898c-70ebb0ae3959","order_by":2,"name":"Delara P Esfarjani","email":"","orcid":"","institution":"Stanford University","correspondingAuthor":false,"prefix":"","firstName":"Delara","middleName":"P","lastName":"Esfarjani","suffix":""},{"id":360679196,"identity":"d092867f-1b70-4bf7-8f44-08527ef43bd9","order_by":3,"name":"Bhavesh Patel","email":"","orcid":"","institution":"FAIR Data Innovations Hub, California Medical Innovations Institute","correspondingAuthor":false,"prefix":"","firstName":"Bhavesh","middleName":"","lastName":"Patel","suffix":""},{"id":360679198,"identity":"7c13bda4-2d3b-4e89-add3-71bad50c9786","order_by":4,"name":"Jennifer Li Pook Than","email":"","orcid":"","institution":"Stanford University","correspondingAuthor":false,"prefix":"","firstName":"Jennifer","middleName":"Li Pook","lastName":"Than","suffix":""},{"id":360679200,"identity":"61b89405-4d0b-4483-9994-93b8d1044a9c","order_by":5,"name":"Aaron Y. Lee","email":"","orcid":"","institution":"University of Washington","correspondingAuthor":false,"prefix":"","firstName":"Aaron","middleName":"Y.","lastName":"Lee","suffix":""},{"id":360679201,"identity":"2b0c783e-5b20-43c4-90c9-31cd943b256f","order_by":6,"name":"Camille Nebeker","email":"","orcid":"","institution":"Herbert Wertheim School of Public Health and Human Longevity Science, UC San Diego, La Jolla, CA","correspondingAuthor":false,"prefix":"","firstName":"Camille","middleName":"","lastName":"Nebeker","suffix":""},{"id":360679202,"identity":"b7dfc208-39a7-4f93-a2ba-dd5db26f2812","order_by":7,"name":"Michael Snyder","email":"","orcid":"","institution":"Stanford University","correspondingAuthor":false,"prefix":"","firstName":"Michael","middleName":"","lastName":"Snyder","suffix":""},{"id":360679203,"identity":"61a4e98c-81cc-4eb5-8a1f-6dc8103f84ed","order_by":8,"name":"Amir Bahmani","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5UlEQVRIiWNgGAWjYDACCQY2ECUH5lQwMDA2MDAYEKXFGEydIUVLYgPRWvhnNz97dKPmTvr8+c3PHhyouCfbwN68TQKvJXeOmRvnHHuWu+EYm7nBgTPFxg08x8rwamG4kWAmncN2OHcDG4OZ9Me2hMQGiRwzvFrkb6R/k875dzhdvo39m8RBkBb5N/i1GNzIMZPObTucwHCMxwyiRYIHvxbDO2fKpHP7DhtuOJZTJnHgTIJxG09asQU+LXK327dJ53w7LC/ffHybxIGKBNl+9sMbb+DTggnYSFM+CkbBKBgFowAbAADT/E27Vu5tFAAAAABJRU5ErkJggg==","orcid":"","institution":"Stanford University","correspondingAuthor":true,"prefix":"","firstName":"Amir","middleName":"","lastName":"Bahmani","suffix":""}],"badges":[],"createdAt":"2024-08-23 19:12:35","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4966049/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4966049/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":66927855,"identity":"b97723d4-8c0b-41e6-bab1-c89c398decc2","added_by":"auto","created_at":"2024-10-18 06:24:41","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":580857,"visible":true,"origin":"","legend":"\u003cp\u003eDifferent types of contexts and prompting settings. Three prompting settings: zero-shot, few-shot, and chain-of-thought.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4966049/v1/16074f9e28559e38ed961b6c.png"},{"id":66927857,"identity":"7bb9e53c-35b8-4201-a3ca-f16140ff237b","added_by":"auto","created_at":"2024-10-18 06:24:41","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2188478,"visible":true,"origin":"","legend":"\u003cp\u003eOverview of LLMs performance. For the various tasks, we employ the following\u003c/p\u003e\n\u003cp\u003emetrics: for \"diabetes detection\" and \"high glucose detection,\" we use the F1-score; for the \"glucose correlation\" task, we utilize three types of correlations: Pearson, Spearman, and cross-correlation; and for the \"age prediction\" task, we use the Mean Absolute Error (MAE). The aggregate performance score for all LLMs, calculated using the optimal prompting settings for each model is as follows: GPT-3.5: 64.6%, GPT-4: 80.7%, GPT-4o: 95.5%, Gemini: 67.7%, Gemini 1.5 pro: 79.9%, and Claude 3 Sonnet: 66.2%.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4966049/v1/ada5e42e50c96b06a2404c5f.png"},{"id":66927856,"identity":"350aa34d-5ced-4d31-9528-1ef4498bd24b","added_by":"auto","created_at":"2024-10-18 06:24:41","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":961477,"visible":true,"origin":"","legend":"\u003cp\u003eNew diabetes insights. (A) Glucose level distribution for diabetic vs. non-diabetic participants. (B) Glucose vs. stress in diabetic and non-diabetic participants with respect to three categories of glucose level (low, normal, and high) and four categories of stress (resting, low, medium, and high). (C) Glucose vs. respiratory rate in diabetic and non-diabetic participants with respect to three categories of glucose level (low, normal, and high) and three categories of respiratory rate (normal, bradypnea, and tachypnea).\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4966049/v1/b4ccda8aad98e97882135020.png"},{"id":70899549,"identity":"032cca7c-3d29-4184-ab93-5704cd2682f8","added_by":"auto","created_at":"2024-12-09 05:32:53","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1994740,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4966049/v1_covered_c7b17c95-3406-4cf5-a76e-014c3894d357.pdf"}],"financialInterests":"Competing interest reported. Dr. Snyder is a cofounder and scientific advisor of Personalis, SensOmics, Qbio, January AI, Fodsel, Filtricine, Protos, RTHM, Iollo, Marble Therapeutics, Crosshair Therapeutics, NextThought, and Mirvie. He is also a scientific advisor of Jupiter, Neuvivo, Swaza, Mitrix, Yuvan, TranscribeGlass, and Applied Cognition. Dr. Lee reports grants from Santen, personal fees from Genentech, personal fees from US FDA, personal fees from Johnson and Johnson, personal fees from Boehringer Ingelheim, non-financial support from iCareWorld, grants from Topcon, grants from Carl Zeiss Meditec, personal fees from Gyroscope, non-financial support from Optomed, non-financial support from Heidelberg, non-financial support from Microsoft, grants from Regeneron, grants from Amazon, grants from Meta, outside the submitted work. The other authors declare no competing interests.","formattedTitle":"Harnessing Large Language Models to Uncover Insights in Diabetes Wearable Data","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4966049/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4966049/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Large Language Models (LLMs) have gained significant attention and are increasingly used by researchers. Concurrently, publicly accessible datasets containing individual-level health information are becoming more available. Some of these datasets, such as the recently released Artificial Intelligence Ready and Equitable Atlas for Diabetes Insights (AI-READI) dataset, include individual-level data from digital wearable technologies. The application of LLMs to gain insights about health from wearable sensor data specific to diabetes is underexplored. This study presents a comprehensive evaluation of multiple LLMs, including GPT-3.5, GPT-4, GPT-4o, Gemini, Gemini 1.5 Pro, and Claude 3 Sonnet, on various diabetes research tasks using diverse prompting methods to evaluate their performance and gain new insights into diabetes and glucose dysregulation. Notably, GPT-4o showed promising performance across tasks with a chain-of-thought prompt design (aggregate performance score of 95.5%). Moreover, using this model, we identified new insights from the dataset, such as the heightened sensitivity to stress among diabetic participants during glucose level fluctuations, which underscores the complex interplay between metabolic and psychological factors. These results demonstrate that LLMs can enhance the pace of discovery and also enable automated interpretation of data for users of wearable devices, including both the research team and the individual wearing the device. Meanwhile, we also emphasize the critical limitations, such as privacy and ethical risks and dataset biases, that must be resolved for real-world application in diabetes health settings. This study highlights the potential and challenges of integrating LLMs into diabetes research and, more broadly, wearables, paving the way for future healthcare advancements, particularly in disadvantaged communities.","manuscriptTitle":"Harnessing Large Language Models to Uncover Insights in Diabetes Wearable Data","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-10-18 06:24:37","doi":"10.21203/rs.3.rs-4966049/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"256be670-61b8-4f40-b941-901be4aeda4c","owner":[],"postedDate":"October 18th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":38370448,"name":"Biological sciences/Computational biology and bioinformatics/Machine learning"},{"id":38370449,"name":"Health sciences/Diseases/Endocrine system and metabolic diseases/Diabetes"}],"tags":[],"updatedAt":"2024-12-09T05:24:14+00:00","versionOfRecord":[],"versionCreatedAt":"2024-10-18 06:24:37","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4966049","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4966049","identity":"rs-4966049","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}