Dataset Information

Bias due to two-stage residual-outcome regression analysis in genetic association studies.

ABSTRACT: Association studies of risk factors and complex diseases require careful assessment of potential confounding factors. Two-stage regression analysis, sometimes referred to as residual- or adjusted-outcome analysis, has been increasingly used in association studies of single nucleotide polymorphisms (SNPs) and quantitative traits. In this analysis, first, a residual-outcome is calculated from a regression of the outcome variable on covariates and then the relationship between the adjusted-outcome and the SNP is evaluated by a simple linear regression of the adjusted-outcome on the SNP. In this article, we examine the performance of this two-stage analysis as compared with multiple linear regression (MLR) analysis. Our findings show that when a SNP and a covariate are correlated, the two-stage approach results in biased genotypic effect and loss of power. Bias is always toward the null and increases with the squared-correlation between the SNP and the covariate (). For example, for , 0.1, and 0.5, two-stage analysis results in, respectively, 0, 10, and 50% attenuation in the SNP effect. As expected, MLR was always unbiased. Since individual SNPs often show little or no correlation with covariates, a two-stage analysis is expected to perform as well as MLR in many genetic studies; however, it produces considerably different results from MLR and may lead to incorrect conclusions when independent variables are highly correlated. While a useful alternative to MLR under , the two -stage approach has serious limitations. Its use as a simple substitute for MLR should be avoided.

SUBMITTER: Demissie S

PROVIDER: S-EPMC3201714 | biostudies-literature | 2011 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Bias due to two-stage residual-outcome regression analysis in genetic association studies.

Demissie Serkalem S Cupples L Adrienne LA

Genetic epidemiology 20110718 7

Association studies of risk factors and complex diseases require careful assessment of potential confounding factors. Two-stage regression analysis, sometimes referred to as residual- or adjusted-outcome analysis, has been increasingly used in association studies of single nucleotide polymorphisms (SNPs) and quantitative traits. In this analysis, first, a residual-outcome is calculated from a regression of the outcome variable on covariates and then the relationship between the adjusted-outcome ...[more]

PMID: 21769934

Similar Datasets

Project description:ImportanceAlthough regression is commonly observed in cutaneous melanoma, it is uncertain whether it is associated with patient prognosis.ObjectiveTo determine whether histologically confirmed regression was associated with better or worse survival in patients with primary cutaneous melanoma.Design, setting, and participantsThis cohort study analyzed data from 2 large cohorts of adults (one in the Netherlands and the other in Australia) with histologically proven, stage 1 and 2 primary, invasive cutaneous melanoma with known regression status treated between 2000 and 2014, with median follow-up times of 4.5 and 11.1 years for the Dutch and Australian cohorts, respectively. For the Dutch patients, population-based data from PALGA, the Dutch Pathology Registry, were used, and follow-up data were retrieved from the Netherlands Cancer Registry. For the Australian patients, data from the database of a large, specialized melanoma treatment center were used.Main outcomes and measuresMultivariable Cox proportional hazards analyses were performed per cohort to assess recurrence-free survival (RFS) and overall survival (OS), and subgroup analyses according to Breslow thickness category and melanoma subtype were performed.ResultsA total of 17 271 Dutch patients and 4980 Australian patients were included. In both cohorts, survival outcomes were better for patients with disease regression. For Dutch patients, the hazard ratio (HR) for those with disease regression was 0.55 (95% CI, 0.48-0.63; P < .001) for RFS and 0.87 (95% CI, 0.79-0.96; P = .004) for OS; for the Australian patients, the HR was 0.61 (95% CI, 0.52-0.72; P < .001) for RFS and 0.73 (95% CI, 0.64-0.84; P < .001) for OS. Subgroup analyses showed that the presence of regression improved RFS within thin and intermediate Breslow thickness melanomas in both cohorts. For patients with superficial spreading melanoma (SSM) subtype, regression improved RFS and OS in both cohorts. For Dutch patients with SSM, the HR for those with disease regression was 0.54 (95% CI, 0.46-0.63; P < .001) for RFS and 0.86 (95% CI, 0.76-0.96; P = .009) for OS; for the Australian patients with SSM, the HR was 0.67 (95% CI, 0.52-0.85; P = .001) for RFS and 0.72 (95% CI, 0.59-0.88; P = .001) for OS.Conclusions and relevanceIn 2 large patient cohorts from 2 different continents, regression was a favorable prognostic factor for patients with stage 1 and 2 melanomas, especially in those with thin and intermediate thickness tumors and those with SSM subtype.

Dataset Information

Bias due to two-stage residual-outcome regression analysis in genetic association studies.

Publications

Bias due to two-stage residual-outcome regression analysis in genetic association studies.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure