Dataset Information

A partial linear model in the outcome-dependent sampling setting to evaluate the effect of prenatal PCB exposure on cognitive function in children.

ABSTRACT: Outcome-dependent sampling (ODS) has been widely used in biomedical studies because it is a cost-effective way to improve study efficiency. However, in the setting of a continuous outcome, the representation of the exposure variable has been limited to the framework of linear models, due to the challenge in terms of both theory and computation. Partial linear models (PLM) are a powerful inference tool to nonparametrically model the relation between an outcome and the exposure variable. In this article, we consider a case study of a PLM for data from an ODS design. We propose a semiparametric maximum likelihood method to make inferences with a PLM. We develop the asymptotic properties and conduct simulation studies to show that the proposed ODS estimator can produce a more efficient estimate than that from a traditional simple random sampling design with the same sample size. Using this newly developed method, we were able to explore an open question in epidemiology: whether in utero exposure to background levels of polychlorinated biphenyls (PCBs) is associated with children's intellectual impairment. Our model provides further insights into the relation between low-level PCB exposure and children's cognitive function. The results shed new light on a body of inconsistent epidemiologic findings.

SUBMITTER: Zhou H

PROVIDER: S-EPMC3182522 | biostudies-literature | 2011 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A partial linear model in the outcome-dependent sampling setting to evaluate the effect of prenatal PCB exposure on cognitive function in children.

Zhou Haibo H Qin Guoyou G Longnecker Matthew P MP

Biometrics 20101029 3

Outcome-dependent sampling (ODS) has been widely used in biomedical studies because it is a cost-effective way to improve study efficiency. However, in the setting of a continuous outcome, the representation of the exposure variable has been limited to the framework of linear models, due to the challenge in terms of both theory and computation. Partial linear models (PLM) are a powerful inference tool to nonparametrically model the relation between an outcome and the exposure variable. In this a ...[more]

PMID: 21039397

Similar Datasets

Project description:ImportanceAlthough the effects of lead (Pb) exposure on neurocognition in children have been confirmed, the individual associations of prenatal Pb exposure and its interaction with genetic factors on cognitive developmental delay (CDD) in children remain unclear.ObjectiveTo investigate the association of prenatal Pb exposure and its interaction with genetic factors with CDD risk.Design, setting, and participantsWomen in Wuhan, China, who had an expected delivery date between March 2014 and December 2017, were recruited for this prospective cohort study. Children were assessed for cognitive development at approximately 2 years of age (March 2016 to December 2019). Maternal venous blood, cord blood, and venous blood from children were collected in a longitudinal follow-up. Data analysis was performed from March 2022 to February 2023.ExposurePrenatal Pb exposure, and genetic risk for cognitive ability evaluated by polygenic risk score constructed with 58 genetic variations.Main outcomes and measuresCognitive developmental delay of children aged approximately 2 years was assessed using the Chinese revision of the Bayley Scale of Infant Development. A series of multivariable logistic regressions was estimated to determine associations between prenatal Pb exposure and CDD among children with various genetic backgrounds, adjusting for confounding variables.ResultsThis analysis included 2361 eligible mother-child pairs (1240 boys [52.5%] and 1121 girls [47.5%]; mean [SD] ages of mothers and children, 28.9 [3.6] years and 24.8 [1.0] months, respectively), with 292 children (12.4%) having CDD. Higher maternal Pb levels were significantly associated with increased risk of CDD (highest vs lowest tertile: odds ratio, 1.55; 95% CI, 1.13-2.13), adjusting for demographic confounders. The association of CDD with maternal Pb levels was more evident among children with higher genetic risk (highest vs lowest tertile: odds ratio, 2.59; 95% CI, 1.48-4.55), adjusting for demographic confounders.Conclusions and relevanceIn this cohort study, prenatal Pb exposure was associated with an increased risk of CDD in children, especially in those with a high genetic risk. These findings suggest that prenatal Pb exposure and genetic background may jointly contribute to an increased risk of CDD for children and indicate the possibility for an integrated strategy to assess CDD risk and improve children's cognitive ability.

Project description:BackgroundBecause fine particulate matter [PM, with aerodynamic diameter ≤2.5μm (PM2.5)] is a ubiquitous environmental exposure, small changes in cognition associated with PM2.5 exposure could have great societal costs. Prior studies have demonstrated a relationship between in utero PM2.5 exposure and cognitive development in urban populations, but it is not known whether these effects are similar in rural populations and whether they persist into late childhood.ObjectivesIn this study, we tested for associations between prenatal PM2.5 exposure and both full-scale and subscale measures of IQ among a longitudinal cohort at age 10.5 y.MethodsThis analysis used data from 568 children enrolled in the Center for the Health Assessment of Mothers and Children of Salinas (CHAMACOS), a birth cohort study in California's agricultural Salinas Valley. Exposures were estimated at residential addresses during pregnancy using state of the art, modeled PM2.5 surfaces. IQ testing was performed by bilingual psychometricians in the dominant language of the child.ResultsA 3-μg/m3 higher average PM2.5 over pregnancy was associated with -1.79 full-scale IQ points [95% confidence interval (CI): -2.98, -0.58], with decrements specifically in Working Memory IQ (WMIQ) and Processing Speed IQ (PSIQ) subscales [WMIQ -1.72 (95% CI: -2.98, -0.45) and PSIQ -1.19 (95% CI: -2.54, 0.16)]. Flexible modeling over the course of pregnancy illustrated mid-to-late pregnancy (months 5-7) as particularly susceptible times, with sex differences in the timing of susceptible windows and in which subscales were most affected [Verbal Comprehension IQ (VCIQ) and WMIQ in males; and PSIQ in females].DiscussionWe found that small increases in outdoor PM2.5 exposure in utero were associated with slightly lower IQ in late childhood, robust to many sensitivity analyses. In this cohort there was a larger effect of PM2.5 on childhood IQ than has previously been observed, perhaps due to differences in PM composition or because developmental disruption could alter the cognitive trajectory and thus appear more pronounced as children get older. https://doi.org/10.1289/EHP10812.

Dataset Information

A partial linear model in the outcome-dependent sampling setting to evaluate the effect of prenatal PCB exposure on cognitive function in children.

Publications

A partial linear model in the outcome-dependent sampling setting to evaluate the effect of prenatal PCB exposure on cognitive function in children.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets