Academic Journal
Προβολή σε EDS

Scalable Gaussian process regression via median posterior inference for estimating the health effects of an environmental mixture.

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: Scalable Gaussian process regression via median posterior inference for estimating the health effects of an environmental mixture.
Συγγραφείς: Sonabend-W A; Google Research, 1600 Amphitheatre Parkway, Mountain View, CA 94043, United States., Zhang J; Department of Statistics, University of California, Davis, One Shields Avenue, Davis, CA 95616, United States., Castro E; Department of Environmental Health, Harvard University, 677 Huntington Ave, Boston, MA 02115, United States., Schwartz J; Department of Environmental Health, Harvard University, 677 Huntington Ave, Boston, MA 02115, United States., Coull BA; Department of Biostatistics, Harvard University, 677 Huntington Ave, Boston, MA 02115, United States., Lu J; Department of Biostatistics, Harvard University, 677 Huntington Ave, Boston, MA 02115, United States.
Πηγή: Biometrics [Biometrics] 2026 Apr 09; Vol. 82 (2).
Τύπος έκδοσης: Journal Article
Γλώσσα: English
Στοιχεία περιοδικού: Publisher: Oxford University Press Country of Publication: England NLM ID: 0370625 Publication Model: Print Cited Medium: Internet ISSN: 1541-0420 (Electronic) Linking ISSN: 0006341X NLM ISO Abbreviation: Biometrics Subsets: MEDLINE
Imprint Name(s): Publication: March 2024- : [Oxford] : Oxford University Press
Original Publication: Alexandria Va : Biometric Society
Ιατρικοί όροι (MeSH): Environmental Exposure*/adverse effects , Environmental Exposure*/statistics & numerical data , Environmental Pollutants* , Complex Mixtures*, Air Pollutants/adverse effects ; Air Pollutants/analysis ; Birth Weight/drug effects ; Biometry/methods ; Normal Distribution ; Bayes Theorem ; Humans ; Computer Simulation ; Regression Analysis ; Models, Statistical ; Monte Carlo Method
Περίληψη: Humans are exposed to complex mixtures of environmental pollutants rather than single chemicals, necessitating methods to quantify the health effects of such mixtures. Research on environmental mixtures provides insights into realistic exposure scenarios, informing regulatory policies that better protect public health. However, statistical challenges, including complex correlations among pollutants and nonlinear multivariate exposure-response relationships, complicate such analyses. A popular Bayesian semi-parametric Gaussian process regression framework addresses these challenges by modeling exposure-response functions with Gaussian processes and performing feature selection to manage high-dimensional exposures while accounting for confounders. Originally designed for small to moderate-sized cohort studies, this framework does not scale well to massive datasets. To address this, we propose a divide-and-conquer strategy, partitioning data, computing posterior distributions in parallel, and combining results using the generalized median. While we focus on Gaussian process models for environmental mixtures, the proposed distributed computing strategy is broadly applicable to other Bayesian models with computationally prohibitive full-sample Markov Chain Monte Carlo fitting. We apply this method to estimate associations between a mixture of ambient air pollutants and 650 000 birthweights recorded in Massachusetts during 2001-2012. Our results reveal negative associations between birthweight and traffic pollution markers, including elemental and organic carbon and PM$_{2.5}$, and positive associations with ozone and vegetation greenness.
(© The Author(s) 2026. Published by Oxford University Press on behalf of The International Biometric Society. All rights reserved. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site-for further information please contact journals.permissions@oup.com.)
References: Environ Res. 2012 Aug;117:100-11. (PMID: 22726801)
Biostatistics. 2015 Jul;16(3):493-508. (PMID: 25532525)
Environ Epidemiol. 2019 Jun 19;3(3):e047. (PMID: 31342007)
Am J Epidemiol. 2019 Sep 1;188(9):1595-1604. (PMID: 31241127)
Neural Netw. 2015 Jan;61:85-117. (PMID: 25462637)
Int J Environ Res Public Health. 2022 Jan 26;19(3):. (PMID: 35162394)
Ann Epidemiol. 2012 Feb;22(2):126-41. (PMID: 22226033)
Environ Pollut. 2022 Aug 1;306:119356. (PMID: 35487468)
Epidemiology. 2019 Sep;30(5):617-623. (PMID: 31386643)
Res Rep Health Eff Inst. 2015 Jun;(183 Pt 1-2):5-50. (PMID: 26333238)
Biometrics. 2007 Dec;63(4):1079-88. (PMID: 18078480)
Grant Information: P30 ES000002 United States ES NIEHS NIH HHS; R01ES032418 United States NH NIH HHS; R01 ES032418 United States ES NIEHS NIH HHS; R01 ES035735 United States ES NIEHS NIH HHS; R01ES035735 United States NH NIH HHS; DMS-2434664 National Science Foundation
Contributed Indexing: Keywords: BKMR; median posterior; multi-pollutant mixtures; scalable Bayesian inference; semi-parametric regression
Substance Nomenclature: 0 (Air Pollutants)
0 (Environmental Pollutants)
0 (Complex Mixtures)
Entry Date(s): Date Created: 20260512 Date Completed: 20260716 Latest Revision: 20260813
Update Code: 20260813
PubMed Central ID: PMC13164939
DOI: 10.1093/biomtc/ujag071
PMID: 42118800
Βάση Δεδομένων: MEDLINE
FullText Text:
  Availability: 0
CustomLinks:
  – Url: http://www.jstor.org/openurl?issn=15410420&date=2026&volume=82&issue=2&spage=
    Name: JSTOR (all content) (s7799221)
    Category: fullText
    Text:
    Icon: https://imageserver.ebscohost.com/branding/jstor/iconJSTOR.gif
    MouseOverText: Full Text from JSTOR
  – Url: https://resolver.ebsco.com/c/fiv2js/result?sid=EBSCO:cmedm&genre=article&issn=15410420&ISBN=&volume=82&issue=2&date=20260409&spage=&pages=&title=Biometrics&atitle=Scalable%20Gaussian%20process%20regression%20via%20median%20posterior%20inference%20for%20estimating%20the%20health%20effects%20of%20an%20environmental%20mixture.&aulast=Sonabend-W%20A&id=DOI:10.1093/biomtc/ujag071
    Name: Full Text Finder (for New FTF UI) (ns324271)
    Category: fullText
    Text: Full Text Finder
    MouseOverText: Full Text Finder
Header DbId: cmedm
DbLabel: MEDLINE
An: 42118800
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Scalable Gaussian process regression via median posterior inference for estimating the health effects of an environmental mixture.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AU" term="%22Sonabend-W+A%22">Sonabend-W A</searchLink>; Google Research, 1600 Amphitheatre Parkway, Mountain View, CA 94043, United States.<br /><searchLink fieldCode="AU" term="%22Zhang+J%22">Zhang J</searchLink>; Department of Statistics, University of California, Davis, One Shields Avenue, Davis, CA 95616, United States.<br /><searchLink fieldCode="AU" term="%22Castro+E%22">Castro E</searchLink>; Department of Environmental Health, Harvard University, 677 Huntington Ave, Boston, MA 02115, United States.<br /><searchLink fieldCode="AU" term="%22Schwartz+J%22">Schwartz J</searchLink>; Department of Environmental Health, Harvard University, 677 Huntington Ave, Boston, MA 02115, United States.<br /><searchLink fieldCode="AU" term="%22Coull+BA%22">Coull BA</searchLink>; Department of Biostatistics, Harvard University, 677 Huntington Ave, Boston, MA 02115, United States.<br /><searchLink fieldCode="AU" term="%22Lu+J%22">Lu J</searchLink>; Department of Biostatistics, Harvard University, 677 Huntington Ave, Boston, MA 02115, United States.
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%220370625%22">Biometrics</searchLink> [Biometrics] 2026 Apr 09; Vol. 82 (2).
– Name: TypePub
  Label: Publication Type
  Group: TypPub
  Data: Journal Article
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: TitleSource
  Label: Journal Info
  Group: Src
  Data: <i>Publisher: </i><searchLink fieldCode="PB" term="%22Oxford+University+Press%22">Oxford University Press </searchLink><i>Country of Publication: </i>England <i>NLM ID: </i>0370625 <i>Publication Model: </i>Print <i>Cited Medium: </i>Internet <i>ISSN: </i>1541-0420 (Electronic) <i>Linking ISSN: </i><searchLink fieldCode="IS" term="%220006341X%22">0006341X </searchLink><i>NLM ISO Abbreviation: </i>Biometrics <i>Subsets: </i>MEDLINE
– Name: PublisherInfo
  Label: Imprint Name(s)
  Group: PubInfo
  Data: <i>Publication</i>: March 2024- : [Oxford] : Oxford University Press<br /><i>Original Publication</i>: Alexandria Va : Biometric Society
– Name: SubjectMESH
  Label: MeSH Terms
  Group: Su
  Data: <searchLink fieldCode="MM" term="%22Environmental+Exposure%22">Environmental Exposure*</searchLink>/<searchLink fieldCode="MM" term="%22Environmental+Exposure+adverse+effects%22">adverse effects</searchLink> <br /><searchLink fieldCode="MM" term="%22Environmental+Exposure%22">Environmental Exposure*</searchLink>/<searchLink fieldCode="MM" term="%22Environmental+Exposure+statistics+%26+numerical+data%22">statistics & numerical data</searchLink> <br /><searchLink fieldCode="MM" term="%22Environmental+Pollutants%22">Environmental Pollutants*</searchLink> <br /><searchLink fieldCode="MM" term="%22Complex+Mixtures%22">Complex Mixtures*</searchLink><br /><searchLink fieldCode="MH" term="%22Air+Pollutants%22">Air Pollutants</searchLink>/<searchLink fieldCode="MH" term="%22Air+Pollutants+adverse+effects%22">adverse effects</searchLink> ; <searchLink fieldCode="MH" term="%22Air+Pollutants%22">Air Pollutants</searchLink>/<searchLink fieldCode="MH" term="%22Air+Pollutants+analysis%22">analysis</searchLink> ; <searchLink fieldCode="MH" term="%22Birth+Weight%22">Birth Weight</searchLink>/<searchLink fieldCode="MH" term="%22Birth+Weight+drug+effects%22">drug effects</searchLink> ; <searchLink fieldCode="MH" term="%22Biometry%22">Biometry</searchLink>/<searchLink fieldCode="MH" term="%22Biometry+methods%22">methods</searchLink> ; <searchLink fieldCode="MH" term="%22Normal+Distribution%22">Normal Distribution</searchLink> ; <searchLink fieldCode="MH" term="%22Bayes+Theorem%22">Bayes Theorem</searchLink> ; <searchLink fieldCode="MH" term="%22Humans%22">Humans</searchLink> ; <searchLink fieldCode="MH" term="%22Computer+Simulation%22">Computer Simulation</searchLink> ; <searchLink fieldCode="MH" term="%22Regression+Analysis%22">Regression Analysis</searchLink> ; <searchLink fieldCode="MH" term="%22Models%2C+Statistical%22">Models, Statistical</searchLink> ; <searchLink fieldCode="MH" term="%22Monte+Carlo+Method%22">Monte Carlo Method</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Humans are exposed to complex mixtures of environmental pollutants rather than single chemicals, necessitating methods to quantify the health effects of such mixtures. Research on environmental mixtures provides insights into realistic exposure scenarios, informing regulatory policies that better protect public health. However, statistical challenges, including complex correlations among pollutants and nonlinear multivariate exposure-response relationships, complicate such analyses. A popular Bayesian semi-parametric Gaussian process regression framework addresses these challenges by modeling exposure-response functions with Gaussian processes and performing feature selection to manage high-dimensional exposures while accounting for confounders. Originally designed for small to moderate-sized cohort studies, this framework does not scale well to massive datasets. To address this, we propose a divide-and-conquer strategy, partitioning data, computing posterior distributions in parallel, and combining results using the generalized median. While we focus on Gaussian process models for environmental mixtures, the proposed distributed computing strategy is broadly applicable to other Bayesian models with computationally prohibitive full-sample Markov Chain Monte Carlo fitting. We apply this method to estimate associations between a mixture of ambient air pollutants and 650 000 birthweights recorded in Massachusetts during 2001-2012. Our results reveal negative associations between birthweight and traffic pollution markers, including elemental and organic carbon and PM$_{2.5}$, and positive associations with ozone and vegetation greenness.<br /> (© The Author(s) 2026. Published by Oxford University Press on behalf of The International Biometric Society. All rights reserved. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site-for further information please contact journals.permissions@oup.com.)
– Name: Ref
  Label: References
  Group: RefInfo
  Data: Environ Res. 2012 Aug;117:100-11. (PMID: <searchLink fieldCode="PM" term="%2222726801%22">22726801)</searchLink><br />Biostatistics. 2015 Jul;16(3):493-508. (PMID: <searchLink fieldCode="PM" term="%2225532525%22">25532525)</searchLink><br />Environ Epidemiol. 2019 Jun 19;3(3):e047. (PMID: <searchLink fieldCode="PM" term="%2231342007%22">31342007)</searchLink><br />Am J Epidemiol. 2019 Sep 1;188(9):1595-1604. (PMID: <searchLink fieldCode="PM" term="%2231241127%22">31241127)</searchLink><br />Neural Netw. 2015 Jan;61:85-117. (PMID: <searchLink fieldCode="PM" term="%2225462637%22">25462637)</searchLink><br />Int J Environ Res Public Health. 2022 Jan 26;19(3):. (PMID: <searchLink fieldCode="PM" term="%2235162394%22">35162394)</searchLink><br />Ann Epidemiol. 2012 Feb;22(2):126-41. (PMID: <searchLink fieldCode="PM" term="%2222226033%22">22226033)</searchLink><br />Environ Pollut. 2022 Aug 1;306:119356. (PMID: <searchLink fieldCode="PM" term="%2235487468%22">35487468)</searchLink><br />Epidemiology. 2019 Sep;30(5):617-623. (PMID: <searchLink fieldCode="PM" term="%2231386643%22">31386643)</searchLink><br />Res Rep Health Eff Inst. 2015 Jun;(183 Pt 1-2):5-50. (PMID: <searchLink fieldCode="PM" term="%2226333238%22">26333238)</searchLink><br />Biometrics. 2007 Dec;63(4):1079-88. (PMID: <searchLink fieldCode="PM" term="%2218078480%22">18078480)</searchLink>
– Name: GrantInfo
  Label: Grant Information
  Group: Grant
  Data: P30 ES000002 United States ES NIEHS NIH HHS; R01ES032418 United States NH NIH HHS; R01 ES032418 United States ES NIEHS NIH HHS; R01 ES035735 United States ES NIEHS NIH HHS; R01ES035735 United States NH NIH HHS; DMS-2434664 National Science Foundation
– Name: SubjectMinor
  Label: Contributed Indexing
  Group:
  Data: <i>Keywords: </i>BKMR; median posterior; multi-pollutant mixtures; scalable Bayesian inference; semi-parametric regression
– Name: NumberCAS
  Label: Substance Nomenclature
  Group: ID
  Data: 0 (Air Pollutants)<br />0 (Environmental Pollutants)<br />0 (Complex Mixtures)
– Name: DateEntry
  Label: Entry Date(s)
  Group: Date
  Data: <i>Date Created: </i>20260512 <i>Date Completed: </i>20260716 <i>Latest Revision: </i>20260813
– Name: DateUpdate
  Label: Update Code
  Group: Date
  Data: 20260813
– Name: PubmedCentralID
  Label: PubMed Central ID
  Group: ID
  Data: PMC13164939
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1093/biomtc/ujag071
– Name: AN
  Label: PMID
  Group: ID
  Data: 42118800
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=cmedm&AN=42118800
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1093/biomtc/ujag071
    Languages:
      – Code: eng
        Text: English
    Subjects:
      – SubjectFull: Air Pollutants adverse effects
        Type: general
      – SubjectFull: Air Pollutants analysis
        Type: general
      – SubjectFull: Birth Weight drug effects
        Type: general
      – SubjectFull: Biometry methods
        Type: general
      – SubjectFull: Normal Distribution
        Type: general
      – SubjectFull: Bayes Theorem
        Type: general
      – SubjectFull: Humans
        Type: general
      – SubjectFull: Computer Simulation
        Type: general
      – SubjectFull: Regression Analysis
        Type: general
      – SubjectFull: Models, Statistical
        Type: general
      – SubjectFull: Monte Carlo Method
        Type: general
      – SubjectFull: Environmental Exposure adverse effects
        Type: general
      – SubjectFull: Environmental Exposure statistics & numerical data
        Type: general
      – SubjectFull: Environmental Pollutants
        Type: general
      – SubjectFull: Complex Mixtures
        Type: general
    Titles:
      – TitleFull: Scalable Gaussian process regression via median posterior inference for estimating the health effects of an environmental mixture.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Sonabend-W A
      – PersonEntity:
          Name:
            NameFull: Zhang J
      – PersonEntity:
          Name:
            NameFull: Castro E
      – PersonEntity:
          Name:
            NameFull: Schwartz J
      – PersonEntity:
          Name:
            NameFull: Coull BA
      – PersonEntity:
          Name:
            NameFull: Lu J
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 09
              M: 04
              Text: 2026 Apr 09
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-electronic
              Value: 1541-0420
          Numbering:
            – Type: volume
              Value: 82
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: Biometrics
              Type: main
ResultId 1