Dataset Information

A comparison of confounding adjustment methods with an application to early life determinants of childhood obesity.

ABSTRACT: We implemented six confounding adjustment methods: (1) covariate-adjusted regression, (2) propensity score (PS) regression, (3) PS stratification, (4) PS matching with two calipers, (5) inverse probability weighting and (6) doubly robust estimation to examine the associations between the body mass index (BMI) z-score at 3 years and two separate dichotomous exposure measures: exclusive breastfeeding v. formula only (n=437) and cesarean section v. vaginal delivery (n=1236). Data were drawn from a prospective pre-birth cohort study, Project Viva. The goal is to demonstrate the necessity and usefulness, and approaches for multiple confounding adjustment methods to analyze observational data. Unadjusted (univariate) and covariate-adjusted linear regression associations of breastfeeding with BMI z-score were -0.33 (95% CI -0.53, -0.13) and -0.24 (-0.46, -0.02), respectively. The other approaches resulted in smaller n (204-276) because of poor overlap of covariates, but CIs were of similar width except for inverse probability weighting (75% wider) and PS matching with a wider caliper (76% wider). Point estimates ranged widely, however, from -0.01 to -0.38. For cesarean section, because of better covariate overlap, the covariate-adjusted regression estimate (0.20) was remarkably robust to all adjustment methods, and the widths of the 95% CIs differed less than in the breastfeeding example. Choice of covariate adjustment method can matter. Lack of overlap in covariate structure between exposed and unexposed participants in observational studies can lead to erroneous covariate-adjusted estimates and confidence intervals. We recommend inspecting covariate overlap and using multiple confounding adjustment methods. Similar results bring reassurance. Contradictory results suggest issues with either the data or the analytic method.

SUBMITTER: Li L

PROVIDER: S-EPMC4337023 | biostudies-literature | 2014 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A comparison of confounding adjustment methods with an application to early life determinants of childhood obesity.

Li L L Kleinman K K Gillman M W MW

Journal of developmental origins of health and disease 20140829 6

We implemented six confounding adjustment methods: (1) covariate-adjusted regression, (2) propensity score (PS) regression, (3) PS stratification, (4) PS matching with two calipers, (5) inverse probability weighting and (6) doubly robust estimation to examine the associations between the body mass index (BMI) z-score at 3 years and two separate dichotomous exposure measures: exclusive breastfeeding v. formula only (n=437) and cesarean section v. vaginal delivery (n=1236). Data were drawn from a ...[more]

PMID: 25171142

Similar Datasets

Project description:BACKGROUND:In stroke studies, ordinal logistic regression (OLR) is often used to analyze outcome on the modified Rankin Scale (mRS), whereas the non-parametric Mann-Whitney measure of superiority (MWS) has also been suggested. It is unclear how these perform comparatively when confounding adjustment is warranted. AIMS:Our aim is to quantify the performance of OLR and MWS in different confounding variable settings. METHODS:We set up a simulation study with three different scenarios; (1) dichotomous confounding variables, (2) continuous confounding variables, and (3) confounding variable settings mimicking a study on functional outcome after stroke. We compared adjusted ordinal logistic regression (aOLR) and stratified Mann-Whitney measure of superiority (sMWS), and also used propensity scores to stratify the MWS (psMWS). For comparability, OLR estimates were transformed to a MWS. We report bias, the percentage of runs that produced a point estimate deviating by more than 0.05 points (point estimate variation), and the coverage probability. RESULTS:In scenario 1, there was no bias in both sMWS and aOLR, with similar point estimate variation and coverage probabilities. In scenario 2, sMWS resulted in more bias (0.04 versus 0.00), and higher point estimate variation (41.6% versus 3.3%), whereas coverage probabilities were similar. In scenario 3, there was no bias in both methods, point estimate variation was higher in the sMWS (6.7%) versus aOLR (1.1%), and coverage probabilities were 0.98 (sMWS) versus 0.95 (aOLR). With psMWS, bias remained 0.00, with less point estimate variation (1.5%) and a coverage probability of 0.95. CONCLUSIONS:The bias of both adjustment methods was similar in our stroke simulation scenario, and the higher point estimate variation in the MWS improved with propensity score based stratification. The stratified MWS is a valid alternative for adjusted OLR only when the ratio of number of strata versus number of observations is relatively low, but propensity score based stratification extends the application range of the MWS.

Project description:Obesity represents a major health burden to both developed and developing countries. Furthermore, the incidence of obesity is increasing in children. Obesity contributes substantially to mortality in the United States by increasing the risk for type 2 diabetes, cardiovascular-related diseases, and other comorbidities. Despite environmental changes over past decades, including increases in high-calorie foods and sedentary lifestyles, there is very clear evidence of a genetic predisposition to obesity risk. Childhood obesity cases can be categorized in one of two ways: syndromic or non-syndromic. Syndromic obesity includes disorders such as Prader-Willi syndrome, Bardet-Biedl syndrome, and Alström syndrome. Non-syndromic cases of obesity can be further separated into rarer instances of monogenic obesity and much more common forms of polygenic obesity. The advent of genome-wide association studies (GWAS) and next-generation sequencing has driven significant advances in our understanding of the genetic contribution to childhood obesity. Many rare and common genetic variants have been shown to contribute to the heritability in obesity, although the molecular mechanisms underlying most of these variants remain unclear. An important caveat of GWAS efforts is that they do not strictly represent gene target discoveries, rather simply the uncovering of robust genetic signals. One clear example of this is with progress in understanding the key obesity signal harbored within an intronic region of the FTO gene. It has been shown that the non-coding region in which the variant actually resides in fact influences the expression of genes distal to FTO instead, specifically IRX3 and IRX5. Such discoveries suggest that associated non-coding variants can be embedded within or next to one gene, but commonly influence the expression of other, more distal effector genes. Advances in genetics and genomics are therefore contributing to a deeper understanding of childhood obesity, allowing for development of clinical tools and therapeutic agents.

Project description:BACKGROUND:Chemical and nonchemical environmental exposures are increasingly suspected to influence the development of obesity, especially during early life, but studies mostly consider single exposure groups. OBJECTIVES:Our study aimed to systematically assess the association between a wide array of early-life environmental exposures and childhood obesity, using an exposome-wide approach. METHODS:The HELIX (Human Early Life Exposome) study measured child body mass index (BMI), waist circumference, skinfold thickness, and body fat mass in 1,301 children from six European birth cohorts age 6-11 y. We estimated 77 prenatal exposures and 96 childhood exposures (cross-sectionally), including indoor and outdoor air pollutants, built environment, green spaces, tobacco smoking, and biomarkers of chemical pollutants (persistent organic pollutants, metals, phthalates, phenols, and pesticides). We used an exposure-wide association study (ExWAS) to screen all exposure-outcome associations independently and used the deletion-substitution-addition (DSA) variable selection algorithm to build a final multiexposure model. RESULTS:The prevalence of overweight and obesity combined was 28.8%. Maternal smoking was the only prenatal exposure variable associated with higher child BMI (z-score increase of 0.28, 95% confidence interval: 0.09, 0.48, for active vs. no smoking). For childhood exposures, the multiexposure model identified particulate and nitrogen dioxide air pollution inside the home, urine cotinine levels indicative of secondhand smoke exposure, and residence in more densely populated areas and in areas with fewer facilities to be associated with increased child BMI. Child blood levels of copper and cesium were associated with higher BMI, and levels of organochlorine pollutants, cobalt, and molybdenum were associated with lower BMI. Similar results were found for the other adiposity outcomes. DISCUSSION:This first comprehensive and systematic analysis of many suspected environmental obesogens strengthens evidence for an association of smoking, air pollution exposure, and characteristics of the built environment with childhood obesity risk. Cross-sectional biomarker results may suffer from reverse causality bias, whereby obesity status influenced the biomarker concentration. https://doi.org/10.1289/EHP5975.

Project description:ObjectivesTo adjust for confounding in observational data, researchers use propensity score matching (PSM), but more advanced methods might be required when dealing with longitudinal data and time-varying treatments as PSM might not include possible changes that occurred over time. This study aims to explore which confounding adjustment methods have been used in longitudinal observational data to estimate a treatment effect and identify potential inappropriate use of PSM.DesignMapping review.Data sourcesWe searched PubMed, from inception up to January 2021, for studies in which a treatment was evaluated using longitudinal observational data.Eligibility criteriaMethodological, non-medical and cost-effectiveness papers were excluded, as were non-English studies and studies that did not study a treatment effect.Data extraction and synthesisStudies were categorised based on time of treatment: at baseline (interventions performed at start of follow-up) or time-varying (interventions received asynchronously during follow-up) and sorted based on publication year, time of treatment and confounding adjustment method. Cumulative time series plots were used to investigate the use of different methods over time. No risk-of-bias assessment was performed as it was not applicable.ResultsIn total, 764 studies were included that met the eligibility criteria. PSM (165/201, 82%) and inverse probability weighting (IPW; 154/502, 31%) were most common for studies with a treatment at baseline (n=201) and time-varying treatment (n=502), respectively. Of the 502 studies with a time-varying treatment, 123 (25%) used PSM with baseline covariates, which might be inappropriate. In the past 5 years, the proportion of studies with a time-varying treatment that used PSM over IPW increased.ConclusionsPSM is the most frequently used method to correct for confounding in longitudinal observational data. In studies with a time-varying treatment, PSM was potentially inappropriately used in 25% of studies. Confounding adjustment methods designed to deal with a time-varying treatment and time-varying confounding are available, but were only used in 45% of the studies with a time-varying treatment.

Dataset Information

A comparison of confounding adjustment methods with an application to early life determinants of childhood obesity.

Publications

A comparison of confounding adjustment methods with an application to early life determinants of childhood obesity.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets