A machine learning-powered early-warning system integrating multi-pollutant exposure and host susceptibility for hospital-acquired acute kidney injury prevention

A machine learning-powered early-warning system integrating multi-pollutant exposure and host susceptibility for hospital-acquired acute kidney injury prevention | 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 Research Article A machine learning-powered early-warning system integrating multi-pollutant exposure and host susceptibility for hospital-acquired acute kidney injury prevention Xiang Yu, XinYan Gong, BaoLong Wang, YuWei Ji, RiLiGe Wu, WanLing Wang, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7144359/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 Background To explore the effects of synergistic exposure to multiple air pollutants on hospitalization-associated acute kidney injury (HA-AKI) and to construct a kidney health index (KHI) that integrates environmental exposure characteristics and host risk factors in order to provide an evidence-based tool for HA-AKI prevention and control. Methods This time-stratified case‒crossover study assessed the acute exposure effects of pollutants such as PM2.5 and CO by analyzing data from 5,672 HA-AKI patients in a tertiary hospital in Beijing from 2014 to 2020. Distributed lag nonlinear modeling (DLNM) was used to analyze lagged effects, a synergistic index (SI) was used to quantify multi-pollutant interactions, and a KHI was constructed by integrating environmental exposure characteristics and host risk factors via XGBoost machine learning. Results Short-term exposure to PM2.5 and CO significantly increased the odds of HA-AKI, with each 10-µg/m³ and 1-mg/m³ increase in these pollutants associated with 15% higher odds of HA-AKI (OR = 1.15, 95% CI: 1.08–1.23) and 18% higher odds (OR = 1.18, 95% CI: 1.09–1.28), respectively. The strongest effect was observed for PM2.5 with a 1-day lag. Synergistic exposure analysis revealed that concurrent high exposure to PM₂.₅ and CO (>75th percentile concentrations) was associated with 52% higher odds of HA-AKI (OR = 1.52, 95% CI: 1.22–1.89; SI = 1.38, p = 0.001) versus low exposure levels. Intensive care unit (ICU) patients had 16% higher PM₂.₅-associated odds than general ward patients (OR = 1.33 vs 1.15, Pinteraction = 0.018), while stage 3 AKI patients showed 23% higher odds than stage 1 AKI patients (OR = 1.38 vs 1.12, Pinteraction = 0.041). Based on these findings, the KHI, integrating PM2.5 (Lag 1), CO (Lag 0), baseline eGFR (< 60 mL/min/1.73 m²), NO2 (Lag 2), and age (≥ 65 years) via XGBoost weighting, demonstrated excellent predictive performance in the validation set (AUC = 0.83, 95% CI: 0.79–0.87), with a sensitivity of 79.3% for a high risk threshold (>80 points). Conclusions Synergistic exposure to PM2.5 and CO is an important environmental trigger for HA-AKI, especially for critically ill patients and those with renal impairment. The KHI provides a clinical tool for assessing the risk of environment‒host interactions by integrating multi-pollutant kinetics with individual susceptibility characteristics, thereby supporting precise decision-making regarding HA-AKI prevention. Trial registration: Not applicable. This study did not involve human subjects or prospective data collection as part of a clinical trial. acute kidney injury air pollution case-crossover design synergistic exposure machine learning health indices Full Text Additional Declarations No competing interests reported. Supplementary Files AdditionalTable1.docx Additional Table 1. Distribution characteristics of ambient air pollutant concentrations. AdditionalTable2.docx Additional Table 2. Synergistic effects of multi-pollutant exposures on HA-AKI risk. SupplementaryFigure1.tif Additional Fig 1. Synergistic Interactions Among Air Pollutants Heatmap of Multi-Pollutant Combination Effects. Note:This matrix illustrates Synergy Index (SI) values for paired and multi-pollutant interactions, with color intensity (purple→yellow-green) quantitatively reflecting transition from additive (SI≈1.0) to strongly synergistic effects (SI→1.5). Matrix coordinates systematically map each interaction: x-axis denotes reference pollutants for individual exposure contrast, while y-axis lists specific combinations (e.g., CO+NO2 intersecting PM2.5 indicates SI=1.38 for PM2.5 as reference). Asterisks (*) denote statistically significant interactions (p<0.05). 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-7144359","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":498658086,"identity":"10feff57-e583-4b1d-91b5-0806e0b0eff5","order_by":0,"name":"Xiang Yu","email":"","orcid":"","institution":"the First Medical Center of Chinese PLA General Hospital, Chinese PLA Institute of Nephrology, National Clinical Research Center for Kidney Diseases","correspondingAuthor":false,"prefix":"","firstName":"Xiang","middleName":"","lastName":"Yu","suffix":""},{"id":498658088,"identity":"3a3f53a2-ba7c-44a2-806f-32c96b255b1a","order_by":1,"name":"XinYan Gong","email":"","orcid":"","institution":"the First Medical Center of Chinese PLA General Hospital, Chinese PLA Institute of Nephrology, National Clinical Research Center for Kidney Diseases","correspondingAuthor":false,"prefix":"","firstName":"XinYan","middleName":"","lastName":"Gong","suffix":""},{"id":498658090,"identity":"1c46d4c1-a87c-4602-be14-2f17ad634e56","order_by":2,"name":"BaoLong Wang","email":"","orcid":"","institution":"the First Medical Center of Chinese PLA General Hospital, Chinese PLA Institute of Nephrology, National Clinical Research Center for Kidney Diseases","correspondingAuthor":false,"prefix":"","firstName":"BaoLong","middleName":"","lastName":"Wang","suffix":""},{"id":498658092,"identity":"30e3ba3e-277b-45c8-b1c5-ce4009824fa7","order_by":3,"name":"YuWei Ji","email":"","orcid":"","institution":"the First Medical Center of Chinese PLA General Hospital, Chinese PLA Institute of Nephrology, National Clinical Research Center for Kidney Diseases","correspondingAuthor":false,"prefix":"","firstName":"YuWei","middleName":"","lastName":"Ji","suffix":""},{"id":498658095,"identity":"9ac01129-da3d-4374-a2ef-32657ac997f7","order_by":4,"name":"RiLiGe Wu","email":"","orcid":"","institution":"Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"RiLiGe","middleName":"","lastName":"Wu","suffix":""},{"id":498658098,"identity":"b3557574-e9e9-4c6e-8a6a-7a0f9b08f5ee","order_by":5,"name":"WanLing Wang","email":"","orcid":"","institution":"Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"WanLing","middleName":"","lastName":"Wang","suffix":""},{"id":498658099,"identity":"dd3fc7e8-e5de-4ff6-8b68-e9e8924fa694","order_by":6,"name":"Sha Luo","email":"","orcid":"","institution":"the First Medical Center of Chinese PLA General Hospital, Chinese PLA Institute of Nephrology, National Clinical Research Center for Kidney Diseases","correspondingAuthor":false,"prefix":"","firstName":"Sha","middleName":"","lastName":"Luo","suffix":""},{"id":498658100,"identity":"f4c5fd81-abc0-4e92-8ee6-75c7068d3478","order_by":7,"name":"MengJie Huang","email":"","orcid":"","institution":"the First Medical Center of Chinese PLA General Hospital, Chinese PLA Institute of Nephrology, National Clinical Research Center for Kidney Diseases","correspondingAuthor":false,"prefix":"","firstName":"MengJie","middleName":"","lastName":"Huang","suffix":""},{"id":498658101,"identity":"52db9c5a-6e3c-4224-9252-34ed141e14ef","order_by":8,"name":"SaSa Nie","email":"","orcid":"","institution":"the First Medical Center of Chinese PLA General Hospital, Chinese PLA Institute of Nephrology, National Clinical Research Center for Kidney Diseases","correspondingAuthor":false,"prefix":"","firstName":"SaSa","middleName":"","lastName":"Nie","suffix":""},{"id":498658102,"identity":"05d1cf75-ad0a-4ed6-b1f5-2f0db4007f03","order_by":9,"name":"GuangYan Cai","email":"","orcid":"","institution":"the First Medical Center of Chinese PLA General Hospital, Chinese PLA Institute of Nephrology, National Clinical Research Center for Kidney Diseases","correspondingAuthor":false,"prefix":"","firstName":"GuangYan","middleName":"","lastName":"Cai","suffix":""},{"id":498658105,"identity":"e635456a-0a3f-4f69-877a-23b3aa138af4","order_by":10,"name":"XiangMei Chen","email":"","orcid":"","institution":"the First Medical Center of Chinese PLA General Hospital, Chinese PLA Institute of Nephrology, National Clinical Research Center for Kidney Diseases","correspondingAuthor":false,"prefix":"","firstName":"XiangMei","middleName":"","lastName":"Chen","suffix":""},{"id":498658106,"identity":"20a4282c-b8a2-4885-b9de-c4ec03a4d02f","order_by":11,"name":"Zhe Feng","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6UlEQVRIie3QsWrDMBCA4TMHynK1VwVKMvQFNDlkSV/lhCBTCp2KhgyBBnkI2dO38FQ6yjhkUnePKX2BdOvWem+x3a2Dvvl+pDuAKPqHxPTde7Zfi9tHrM5s1/1JCkvtLwGNGjmjzuHUn0xgdVM9OUxKCvn4bYsDPgYBagoCZ3KVW70RkBU77k6Sva/JUjo/LB8a/XINMryW3Qmm3L4iERrz3OggQMm7nkSQqq+cSjYN5/fa4YCESLXrc1KGYw7DEim4PbI348IZyeFEvbtMD1hf2PpFhlh9fNr1JCv23ckP9LfxKIqi6FffaShOLPs1iScAAAAASUVORK5CYII=","orcid":"","institution":"the First Medical Center of Chinese PLA General Hospital, Chinese PLA Institute of Nephrology, National Clinical Research Center for Kidney Diseases","correspondingAuthor":true,"prefix":"","firstName":"Zhe","middleName":"","lastName":"Feng","suffix":""}],"badges":[],"createdAt":"2025-07-17 03:08:19","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-7144359/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7144359/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":94824440,"identity":"53504d4a-69ce-44c6-b4dc-3264ab965b31","added_by":"auto","created_at":"2025-10-31 06:49:00","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1140487,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7144359/v1_covered_2a0dde83-bb97-43a4-b639-1535d8a47961.pdf"},{"id":88975053,"identity":"40dd6253-df57-4077-b9e6-dbd1b609b72d","added_by":"auto","created_at":"2025-08-13 10:11:46","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":12656,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional Table 1\u003c/strong\u003e. Distribution characteristics of ambient air pollutant concentrations.\u003c/p\u003e","description":"","filename":"AdditionalTable1.docx","url":"https://assets-eu.researchsquare.com/files/rs-7144359/v1/06cd4979fc691762427c8750.docx"},{"id":88973983,"identity":"b8eda01a-d071-4307-b7a3-101d296bbbb2","added_by":"auto","created_at":"2025-08-13 10:03:46","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":12203,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional Table 2\u003c/strong\u003e. Synergistic effects of multi-pollutant exposures on HA-AKI risk.\u003c/p\u003e","description":"","filename":"AdditionalTable2.docx","url":"https://assets-eu.researchsquare.com/files/rs-7144359/v1/20394e6b57354f8f85af1140.docx"},{"id":88973991,"identity":"f2af55a7-a030-40ca-9fed-58acc0e55d6a","added_by":"auto","created_at":"2025-08-13 10:03:46","extension":"tif","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":5656554,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional Fig 1\u003c/strong\u003e. Synergistic Interactions Among Air Pollutants Heatmap of Multi-Pollutant Combination Effects. Note:This matrix illustrates Synergy Index (SI) values for paired and multi-pollutant interactions, with color intensity (purple→yellow-green) quantitatively reflecting transition from additive (SI≈1.0) to strongly synergistic effects (SI→1.5). Matrix coordinates systematically map each interaction: x-axis denotes reference pollutants for individual exposure contrast, while y-axis lists specific combinations (e.g., CO+NO2 intersecting PM2.5 indicates SI=1.38 for PM2.5 as reference). Asterisks (*) denote statistically significant interactions (p\u0026lt;0.05).\u003c/p\u003e","description":"","filename":"SupplementaryFigure1.tif","url":"https://assets-eu.researchsquare.com/files/rs-7144359/v1/fdd753e0fdcfc075be827584.tif"}],"financialInterests":"No competing interests reported.","formattedTitle":"A machine learning-powered early-warning system integrating multi-pollutant exposure and host susceptibility for hospital-acquired acute kidney injury prevention","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":true,"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":"acute kidney injury, air pollution, case-crossover design, synergistic exposure, machine learning, health indices","lastPublishedDoi":"10.21203/rs.3.rs-7144359/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7144359/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eTo explore the effects of synergistic exposure to multiple air pollutants on hospitalization-associated acute kidney injury (HA-AKI) and to construct a kidney health index (KHI) that integrates environmental exposure characteristics and host risk factors in order to provide an evidence-based tool for HA-AKI prevention and control.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eThis time-stratified case‒crossover study assessed the acute exposure effects of pollutants such as PM2.5 and CO by analyzing data from 5,672 HA-AKI patients in a tertiary hospital in Beijing from 2014 to 2020. Distributed lag nonlinear modeling (DLNM) was used to analyze lagged effects, a synergistic index (SI) was used to quantify multi-pollutant interactions, and a KHI was constructed by integrating environmental exposure characteristics and host risk factors via XGBoost machine learning.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eShort-term exposure to PM2.5 and CO significantly increased the odds of HA-AKI, with each 10-\u0026micro;g/m\u0026sup3; and 1-mg/m\u0026sup3; increase in these pollutants associated with 15% higher odds of HA-AKI (OR\u0026thinsp;=\u0026thinsp;1.15, 95% CI: 1.08\u0026ndash;1.23) and 18% higher odds (OR\u0026thinsp;=\u0026thinsp;1.18, 95% CI: 1.09\u0026ndash;1.28), respectively. The strongest effect was observed for PM2.5 with a 1-day lag. Synergistic exposure analysis revealed that concurrent high exposure to PM₂.₅ and CO (\u0026gt;75th percentile concentrations) was associated with 52% higher odds of HA-AKI (OR\u0026thinsp;=\u0026thinsp;1.52, 95% CI: 1.22\u0026ndash;1.89; SI\u0026thinsp;=\u0026thinsp;1.38, p\u0026thinsp;=\u0026thinsp;0.001) versus low exposure levels. Intensive care unit (ICU) patients had 16% higher PM₂.₅-associated odds than general ward patients (OR\u0026thinsp;=\u0026thinsp;1.33 vs 1.15, Pinteraction\u0026thinsp;=\u0026thinsp;0.018), while stage 3 AKI patients showed 23% higher odds than stage 1 AKI patients (OR\u0026thinsp;=\u0026thinsp;1.38 vs 1.12, Pinteraction\u0026thinsp;=\u0026thinsp;0.041). Based on these findings, the KHI, integrating PM2.5 (Lag 1), CO (Lag 0), baseline eGFR (\u0026lt;\u0026thinsp;60 mL/min/1.73 m\u0026sup2;), NO2 (Lag 2), and age (\u0026ge;\u0026thinsp;65 years) via XGBoost weighting, demonstrated excellent predictive performance in the validation set (AUC\u0026thinsp;=\u0026thinsp;0.83, 95% CI: 0.79\u0026ndash;0.87), with a sensitivity of 79.3% for a high risk threshold (\u0026gt;80 points).\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e\u003cp\u003eSynergistic exposure to PM2.5 and CO is an important environmental trigger for HA-AKI, especially for critically ill patients and those with renal impairment. The KHI provides a clinical tool for assessing the risk of environment‒host interactions by integrating multi-pollutant kinetics with individual susceptibility characteristics, thereby supporting precise decision-making regarding HA-AKI prevention.\u003c/p\u003e\u003ch2\u003eTrial registration:\u003c/h2\u003e\u003cp\u003eNot applicable. This study did not involve human subjects or prospective data collection as part of a clinical trial.\u003c/p\u003e","manuscriptTitle":"A machine learning-powered early-warning system integrating multi-pollutant exposure and host susceptibility for hospital-acquired acute kidney injury prevention","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-13 10:03:41","doi":"10.21203/rs.3.rs-7144359/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":"d34f6ce1-66d0-4172-bedb-644064035306","owner":[],"postedDate":"August 13th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-10-30T16:53:14+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-13 10:03:41","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7144359","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7144359","identity":"rs-7144359","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}