Response to “Comment on ‘A Quantile-Based g-Computation Approach to Addressing the Effects of Exposure Mixtures’”

Keil, Alexander P.; Buckley, Jessie P.; O’Brien, Katie M.; Ferguson, Kelly K.; Zhao, Shanshan; White, Alexandra J.

doi:10.1289/ehp8820

Cited by 7 publications

(7 citation statements)

References 8 publications

(10 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…have since responded to that criticism defending their methods. 27 Similar to the Keil et al. findings when assessing unidirectional simulations, quantile g-computation is more powerful and less biased than the WQSBS_Split models in determining a mixture coefficient.…”

Section: Discussionmentioning

confidence: 56%

A Permutation Test-Based Approach to Strengthening Inference on the Effects of Environmental Mixtures: Comparison between Single-Index Analytic Methods

Day

Sathyanarayana

LeWinn

et al. 2022

Environ Health Perspect

View full text Add to dashboard Cite

Background: Optimization of mixture analyses is critical to assess potential impacts of human exposures to multiple pollutants. Single-index regression methods quantify total mixture association and chemical component contributions. Single-index methods include several variants of quantile g-computation (QGC) and weighted quantile sum regression (WQSr), though each has limitations. Objectives: We developed a novel permutation test for WQSr and compared its performance to extant versions of WQSr and QGC in simulation studies and an analysis of prenatal phthalates and childhood cognition. Methods: WQSr uses ensemble nonlinear optimization to identify weights for mixture exposures in an index associated with the outcome in a prespecified direction, with ensembles based on bootstrap resampling (WQSBS) or random subsetting of exposures (WQSRS). Statistical significance can be assessed without splitting the data (Nosplit), by splitting the data into training and test sets (Split), by repeatedly holding out test sets (RH), or by using a novel permutation test (PT) to obtain a more accurate -value. QGC instead provides a sum mixture coefficient and component coefficients with no constraints on direction. In simulations, we compared false positive rates (FPR) and power to detect true associations and accuracy in estimating mixture weights. We also estimated associations between prenatal phthalate mixtures and childhood IQ in the Conditions Affecting Neurocognitive Development and Learning in Early Childhood cohort using each method. Results: FPR was well controlled at in all but the Nosplit WQSr variants. Among these methods, the WQSBS and WQSRS PT variants had the highest power (89%–98%), with lower power for QGC (85%–93%) and RH (60%–97%) or Split WQSr variants (40%–78%). WQSr methods estimated mixture weights 2–4 times more accurately than the QGC method. Coefficients for mixture associations with full scale IQ varied 3- to 4-fold across analytic methods. Discussion: WQSr paired with our novel permutation test best balanced power and false positive rate when assessing a mixture effect. As new methods develop, each should be examined for performance and applicability. https://doi.org/10.1289/EHP10570

show abstract

Section: Discussionmentioning

confidence: 56%

A Permutation Test-Based Approach to Strengthening Inference on the Effects of Environmental Mixtures: Comparison between Single-Index Analytic Methods

Day

Sathyanarayana

LeWinn

et al. 2022

Environ Health Perspect

View full text Add to dashboard Cite

show abstract

“…We fit four quantile g-computation models with SRS total raw score as the outcome and the 6 phthalate metabolites (i.e., ΣDEHP, MBP, MBzP, MCPP, MEP, and MiBP) as the independent variables. 39,40 Model 1 was a joint marginal structural model given by falsedouble-struckE(YXq|Z,ψ,η)=g(ψ0+ψ1Sq+ηZ) where falseY was SRS total raw score, falseXq was the phthalates decile, falseboldZ were the covariates, falseg(⋅) was a link function in a generalized linear model (e.g., the identity function in the case of a linear regression model for the expected value of falseY), falseψ0 was the model intercept, falseψ1 was the expected change in the outcome, given a one quantile increase in all exposures simultaneously, falseSq is an “index” that represented a joint value of the phthalates, and falseη was a set of model coefficients for the covariates …”

Section: Methodsmentioning

confidence: 99%

“…Next, we calculated the weights of each phthalate within this mixture and defined as the proportion of the overall effects contributed by each phthalate in either the positive or negative direction. [39][40][41] These weights represented the contribution of the individual mixture components to the overall mixture effect. 42,43 Hence, the weight could range from -1 to 1, with all negative weights (i.e., negative partial mixture effects) summed to -1 and all positive weights (i.e., positive partial mixture effects) summed to 1.…”

mentioning

confidence: 99%

A Mixture of Urinary Phthalate Metabolite Concentrations During Pregnancy and Offspring Social Responsiveness Scale Scores

Yu,

Braun,

Lyall

et al. 2023

Epidemiology

View full text Add to dashboard Cite

Background: Phthalates are a group of chemicals with ubiquitous exposure worldwide. Exposures to phthalates during pregnancy may play a role in autism spectrum disorder (ASD) etiology by disrupting hormone levels or directly impacting fetal neurodevelopment. However, there is little research quantifying the aggregate effect of phthalates on child ASD-related behaviors. Methods: We used data from two prospective pregnancy and birth cohorts – the Health Outcomes and Measures of the Environment (HOME) and the Early Autism Risk Longitudinal Investigation (EARLI). HOME is a general population cohort while participants in EARLI were at higher familial risk for ASD. Using quantile g-computation and linear regression models, we assessed the joint and individual associations of a mixture of six phthalate metabolites during pregnancy with child ASD-related traits measured by Social Responsiveness Scale (SRS) scores at ages 3-8 years. Results: Our analyses included 271 participants from HOME and 166 from EARLI. There were imprecise associations between the phthalate mixture and SRS total raw scores in HOME (difference in SRS scores per decile increase in every phthalate = 1.3; 95% confidence interval [CI]: -0.2, 2.8) and EARLI (difference in SRS scores per decile increase in every phthalate = -0.9; 95% CI: -3.5, 1.7). Conclusions: The cohort-specific effect sizes of the pthalates–SRS associations were small and confidence intervals were imprecise. These results suggest that if there are associations between phthalate metabolites during pregnancy and child SRS scores, they may differ across populations with different familial liabilities. Further studies with larger sample sizes are warranted.

show abstract

“…The model without covariates is given by, gcomp differs in that WQS first estimates the weights 𝑤 𝑝 on the training data and then estimates 𝜓 and its p-value on the validation data based on the estimated weights; while Q-gcomp use all the data to estimate 𝛽 𝑝 and obtain 𝜓. We implement Q-gcomp via R package "qgcomp" (version 2.8.5) [47].…”

Section: Q-gcompmentioning

confidence: 99%

Statistical methods for chemical mixtures: a roadmap for practitioners

Hao,

Cathey,

Aung

et al. 2024

Preprint

View full text Add to dashboard Cite

Quantitative characterization of the health impacts associated with exposure to chemical mixtures has received considerable attention in current environmental and epidemiological studies. With many existing statistical methods and emerging approaches, it is important for practitioners to understand when each method is best suited for their inferential goals. In this study, we conduct a review and comparison of 11 analytical methods available for use in mixtures research, through extensive simulation studies for continuous and binary outcomes. These methods fall in three different classes: identifying important components of a mixture, identifying interactions and creating a summary score for risk stratification and prediction. We carry out an illustrative data analysis in the PROTECT birth cohort from Puerto Rico. Most importantly we develop an integrated package "CompMix" that provides a platform for mixtures analysis where the practitioner can implement a pipeline for several types of mixtures analysis. Our simulation results suggest that the choice of methods depends on the goal of analysis and there is no clear winner across the board. For selection of important toxicants in the mixture and for identifying interactions, Elastic net by Zou et al. (Enet), Lasso for Hierarchical Interactions by Bien et al (HierNet), Selection of nonlinear interactions by a forward stepwise algorithm by Narisetty et al. (SNIF) have the most stable performance across simulation settings. Additionally, the predictive performance of the Super Learner ensembling method by Van de Laan et al. and HierNet are found to be superior to the rest of the methods. For overall summary or a cumulative measure, we find that using the Super Learner to combine multiple Environmental Risk Scores can lead to improved risk stratification properties. We have developed an R package "CompMix: A comprehensive toolkit for environmental mixtures analysis", allowing users to implement a variety of tasks under different settings and compare the findings. In summary, our study offers guidelines for selecting appropriate statistical methods for addressing specific scientific questions related to mixtures research. We identify critical gaps where new and better methods are needed.

show abstract

Response to “Comment on ‘A Quantile-Based g-Computation Approach to Addressing the Effects of Exposure Mixtures’”

Cited by 7 publications

References 8 publications

A Permutation Test-Based Approach to Strengthening Inference on the Effects of Environmental Mixtures: Comparison between Single-Index Analytic Methods

A Permutation Test-Based Approach to Strengthening Inference on the Effects of Environmental Mixtures: Comparison between Single-Index Analytic Methods

A Mixture of Urinary Phthalate Metabolite Concentrations During Pregnancy and Offspring Social Responsiveness Scale Scores

Statistical methods for chemical mixtures: a roadmap for practitioners

Contact Info

Product

Resources

About