MétaCan
Menu
Back to cohort
Record W4409349100 · doi:10.1016/j.neuro.2025.04.001

Using machine learning to investigate the influence of the prenatal chemical exposome on neurodevelopment of young children

2025· article· en· W4409349100 on OpenAlexafffund
Gillian England-Mason, Sarah J. MacEachern, Kimberly Amador, Munawar Hussain Soomro, Anthony Reardon, Amy M. MacDonald, David W. Kinniburgh, Nicole Letourneau, Gerald F. Giesbrecht, Jonathan W. Martin, Nils D. Forkert, Deborah Dewey

Bibliographic record

VenueNeuroToxicology · 2025
Typearticle
Languageen
FieldEnvironmental Science
TopicHealth, Environment, Cognitive Aging
Canadian institutionsUniversity of AlbertaUniversity of CalgaryAlberta Children's Hospital
FundersAlberta Children's Hospital Research InstituteNational Institutes of HealthCanada Research ChairsNational Institute of Environmental Health SciencesCanadian Institutes of Health ResearchAlberta Innovates - Health Solutions
KeywordsExposomePrenatal exposureNeuroscienceMedicinePhysiologyPsychologyEnvironmental healthDevelopmental psychologyBiologyComputational biologyPregnancyGeneticsGestation

Abstract

fetched live from OpenAlex

Research investigating the prenatal chemical exposome and child neurodevelopment has typically focused on a limited number of chemical exposures and controlled for sociodemographic factors and maternal mental health. Emerging machine learning approaches may facilitate more comprehensive examinations of the contributions of chemical exposures, sociodemographic factors, and maternal mental health to child neurodevelopment. A machine learning pipeline that utilized feature selection and ranking was applied to investigate which common prenatal chemical exposures and sociodemographic factors best predict neurodevelopmental outcomes in young children. Data from 406 maternal-child pairs enrolled in the APrON study were used. Maternal concentrations of 32 environmental chemical exposures ( i.e., phthalates, bisphenols, per- and polyfluoroalkyl substances (PFAS), metals, trace elements) measured during pregnancy and 11 sociodemographic factors, as well as measures of maternal mental health and urinary creatinine were entered into the machine learning pipeline. The pipeline, which consisted of a RReliefF variable selection algorithm and support vector machine regression model, was used to identify and rank the best subset of variables predictive of cognitive, language, and motor development outcomes on the Bayley Scales of Infant Development-Third Edition (Bayley-III) at 2 years of age. Bayley-III cognitive scores were best predicted using 29 variables, resulting in a correlation coefficient of r = 0.27 (R 2 =0.07). For language outcomes, 45 variables led to the best result (r = 0.30; R 2 =0.09), whereas for motor outcomes 33 variables led to the best result (r = 0.28, R 2 =0.09). Environmental chemicals, sociodemographic factors, and maternal mental health were found to be highly ranked predictors of cognitive, language, and motor development in young children. Our findings demonstrate the potential of machine learning approaches to identify and determine the relative importance of different predictors of child neurodevelopmental outcomes. Future developmental neurotoxicology research should consider the prenatal chemical exposome as well as sample characteristics such as sociodemographic factors and maternal mental health as important predictors of child neurodevelopment. • Machine learning can be used to examine relationships among prenatal environmental chemical exposures and neurodevelopment. • Environmental chemicals were highly ranked predictors of cognitive, language, and motor development. • The highly ranked environmental chemical predictors differ based on the neurodevelopmental outcome assessed. • Sociodemographics factors were ranked as top predictors of cognitive, language, and motor outcomes. • Research needs to incorporate sociodemographics and environmental exposures in predictive models

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.001
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Observational · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.331
Threshold uncertainty score0.427

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0000.001
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.000
Science and technology studies0.0000.001
Scholarly communication0.0000.000
Open science0.0010.001
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.015
GPT teacher head0.258
Teacher spread0.243 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one teacher head, not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designObservational
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations6
Published2025
Admission routes2
Has abstractyes

Explore more

Same venueNeuroToxicologySame topicHealth, Environment, Cognitive AgingFrench-language works237,207