Using a Mixture-of-Bivariate-Regressions Model to Explore Heterogeneity of Effects of the Use of Inhaled Corticosteroids on Gestational Age and Birth Weight Among Pregnant Women With Asthma

Samoilenko, Mariia; Blais, Lucie; Boucoiran, Isabelle; Lefebvre, Geneviève

doi:10.1093/aje/kwy105

Cited by 5 publications

(6 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…43 While interpatient variability in response to treatment with inhaled corticosteroids for the control of asthma is well established, 44,45 it is reasonable to believe that similar heterogeneity can prevail for the effect of inhaled corticosteroids on secondary outcomes such as birthweight and gestational age. 22 In what follows, we further investigate this issue and illustrate our application of the model of Schwartz et al 7 for estimating the natural direct and indirect effects of maternal exposure to inhaled corticosteroids during pregnancy on birthweight when gestational age is the potential mediator of interest.…”

Section: Applicationmentioning

confidence: 99%

“…The univariate mixture of regression model naturally arises from the presence of omitted or unmeasured binary or categorical predictors. To illustrate this and following Samoilenko et al, suppose that birthweight ( Y ) is a function of fetal sex ( S ) and a given genotype ( G , with 0/1 coding) such as

Y = β_{1} + s β_{2} + sans-serifg β_{3} + sans-serifg s β_{4} + ε, ε \sim N false(0, σ^{2} false) .

Model can alternatively be written as

Y = β_{sans-serifg, 1} + s β_{sans-serifg, 2} + ε,

where

β_{sans-serifg, 1} = β_{1} + sans-serifg β_{3}

and

β_{sans-serifg, 2} = β_{2} + sans-serifg β_{4}

. If G is unmeasured, then model is a mixture of two univariate regressions, and assuming that genotype G and fetal sex are independent, the corresponding mixing proportions are given by the prevalence of the levels of G (eg, π 1 = P ( G = 0) and π 2 = P ( G = 1)).…”

Section: Review Of Univariate Finite Mixture Of Regressionsmentioning

confidence: 99%

Section: Applicationmentioning

confidence: 99%

“…[16][17][18][19][20][21] The univariate mixture of regression model (1) naturally arises from the presence of omitted or unmeasured binary or categorical predictors. To illustrate this and following Samoilenko et al, 22 suppose that birthweight (Y) is a function of fetal sex (S) and a given genotype (G, with 0/1 coding) such as…”

Section: Review Of Univariate Finite Mixture Of Regressionsmentioning

confidence: 99%

“…These variables, identified as risk factors for low birthweight or prematurity or both, can be divided in the following four categories: (i) mother's and baby's characteristics; (ii) maternal chronic conditions in the year before or during pregnancy; (iii) pregnancy-related variables; and (iv) asthma-related variables. The complete list of variables can be found in Samoilenko et al 22 Schwartz et al 7 reported results for a three-component mixture based on n = 336 129 observations from the North Carolina Detailed Birth Record database. Although our sample size is much smaller (n = 6197), we also implemented the mixture model with up to three components.…”

Section: Data Cohort and Analysesmentioning

confidence: 99%

See 4 more Smart Citations

A Bayesian finite mixture of bivariate regression model for causal mediation analyses

Lefebvre

Samoilenko

Boucoiran

et al. 2018

Statistics in Medicine

Self Cite

View full text Add to dashboard Cite

Building on the work of Schwartz et al, Joint Bayesian analysis of birthweight and censored gestational age using finite mixture models in Statistics in Medicine, we propose a Bayesian finite mixture of bivariate regression model for causal mediation analyses. Using an identifiability condition within each component of the mixture, we express the natural direct and indirect effects of the exposure on the outcome as functions of the component-specific regression coefficients. On the basis of simulated data, we examine the behavior of the model for estimating these effects in situations where the associations between exposure, mediator, and outcome are confounded or not. Additionally, we demonstrate that this mixture model can be used to account for heterogeneity arising through unmeasured binary or categorical mediator-outcome confounders. Considering gestational age as a potential mediator, we then illustrate our mediation mixture model to estimate the natural direct and indirect effects of exposure to inhaled corticosteroids during pregnancy on birthweight using a cohort of asthmatic women from the province of Quebec (Canada).

show abstract