Hypothesis Tests for Neyman's Bias in Case-Control Studies.
Ontology highlight
ABSTRACT:
SUBMITTER: Swanson DM
PROVIDER: S-EPMC6150474 | biostudies-literature | 2018
REPOSITORIES: biostudies-literature
Publications
Hypothesis Tests for Neyman's Bias in Case-Control Studies.
Swanson D M DM Anderson C D CD Betensky R A RA
Journal of applied statistics 20171116 11
Survival bias is a long-recognized problem in case-control studies, and many varieties of bias can come under this umbrella term. We focus on one of them, termed Neyman's bias or "prevalence-incidence bias." It occurs in case-control studies when exposure affects both disease and disease-induced mortality, and we give a formula for the observed, biased odds ratio under such conditions. We compare our result with previous investigations into this phenomenon and consider models under which this bi ...[more]
Similar Datasets
Project description:Advances in sequencing technology allow assessing the impact of rare variation on common disorders. For this purpose, methods combine rare variants across a gene and compare an aggregate statistic between cases and controls. However, sequencing many individuals is costly. Hence, it is necessary to identify case samples that are most likely to result in powerful tests under realistic model assumptions. Power can be increased by selecting cases that are highly likely to carry risk variants. As rare variants that contribute to the heritability of a disease co-segregate among affected family members, selecting cases that have affected family members may increase the power of rare variant tests considerably. Here I compare sequencing random cases to cases ascertained to have affected family members. I quantify the power of the different approaches and provide criteria for sample selection under different models of inheritance. Under a model of multiplicative gene-gene interaction, a sample of random cases has to be 2-16-fold larger to achieve the same power as a sample of cases ascertained to have affected family members. However, in traits with high heritability this power gain can be reduced or even reversed under models of additive gene-gene interaction. Hence study designs should depend on the studied disease's heritability and on the available sample size. I also show that selecting cases that share both chromosomes identical by descent with an affected sibling at candidate regions can result in a further power gain.
| S-EPMC3449077 | biostudies-literature
Project description:Genome-wide association studies (GWAs) have identified thousands of DNA loci associated with a variety of traits. Statistical inference is almost always based on single marker hypothesis tests of association and the respective p-values with Bonferroni correction. Since commercially available genomic arrays interrogate hundreds of thousands or even millions of loci simultaneously, many causal yet undetected loci are believed to exist because the conditional power to achieve a genome-wide significance level can be low, in particular for markers with small effect sizes and low minor allele frequencies and in studies with modest sample size. However, the correlation between neighboring markers in the human genome due to linkage disequilibrium (LD) resulting in correlated marker test statistics can be incorporated into multi-marker hypothesis tests, thereby increasing power to detect association. Herein, we establish a theoretical benchmark by quantifying the maximum power achievable for multi-marker tests of association in case-control studies, achievable only when the causal marker is known. Using that genotype correlations within an LD block translate into an asymptotically multivariate normal distribution for score test statistics, we develop a set of weights for the markers that maximize the non-centrality parameter, and assess the relative loss of power for other approaches. We find that the method of Conneely and Boehnke (2007) based on the maximum absolute test statistic observed in an LD block is a practical and powerful method in a variety of settings. We also explore the effect on the power that prior biological or functional knowledge used to narrow down the locus of the causal marker can have, and conclude that this prior knowledge has to be very strong and specific for the power to approach the maximum achievable level, or even beat the power observed for methods such as the one proposed by Conneely and Boehnke (2007).
| S-EPMC3863805 | biostudies-other
Project description:To account for between-study heterogeneity in meta-analysis of diagnostic accuracy studies, bivariate random effects models have been recommended to jointly model the sensitivities and specificities. As study design and population vary, the definition of disease status or severity could differ across studies. Consequently, sensitivity and specificity may be correlated with disease prevalence. To account for this dependence, a trivariate random effects model had been proposed. However, the proposed approach can only include cohort studies with information estimating study-specific disease prevalence. In addition, some diagnostic accuracy studies only select a subset of samples to be verified by the reference test. It is known that ignoring unverified subjects may lead to partial verification bias in the estimation of prevalence, sensitivities, and specificities in a single study. However, the impact of this bias on a meta-analysis has not been investigated. In this paper, we propose a novel hybrid Bayesian hierarchical model combining cohort and case-control studies and correcting partial verification bias at the same time. We investigate the performance of the proposed methods through a set of simulation studies. Two case studies on assessing the diagnostic accuracy of gadolinium-enhanced magnetic resonance imaging in detecting lymph node metastases and of adrenal fluorine-18 fluorodeoxyglucose positron emission tomography in characterizing adrenal masses are presented.
| S-EPMC4245380 | biostudies-literature
Project description:In case-control studies, genetic associations for complex diseases may be probed either with single-locus tests or with haplotype-based tests. Although there are different views on the relative merits and preferences of the two test strategies, haplotype-based analyses are generally believed to be more powerful to detect genes with modest effects. However, a main drawback of haplotype-based association tests is the large number of distinct haplotypes, which increases the degrees of freedom for corresponding test statistics and thus reduces the statistical power. To decrease the degrees of freedom and enhance the efficiency and power of haplotype analysis, we propose an improved haplotype clustering method that is based on the haplotype cladistic analysis developed by Durrant et al. In our method, we attempt to combine the strengths of single-locus analysis and haplotype-based analysis into one single test framework. Novel in our method is that we develop a more informative haplotype similarity measurement by using p-values obtained from single-locus association tests to construct a measure of weight, which to some extent incorporates the information of disease outcomes. The weights are then used in computation of similarity measures to construct distance metrics between haplotype pairs in haplotype cladistic analysis. To assess our proposed new method, we performed simulation analyses to compare the relative performances of (1) conventional haplotype-based analysis using original haplotype, (2) single-locus allele-based analysis, (3) original haplotype cladistic analysis (CLADHC) by Durrant et al., and (4) our weighted haplotype cladistic analysis method, under different scenarios. Our weighted cladistic analysis method shows an increased statistical power and robustness, compared with the methods of haplotype cladistic analysis, single-locus test, and the traditional haplotype-based analyses. The real data analyses also show that our proposed method has practical significance in the human genetics field.
| S-EPMC1829402 | biostudies-literature
Project description:Epidemiological studies often utilize stratified data in which rare outcomes or exposures are artificially enriched. This design can increase precision in association tests but distorts predictions when applying classifiers on nonstratified data. Several methods correct for this so-called sample selection bias, but their performance remains unclear especially for machine learning classifiers. With an emphasis on two-phase case-control studies, we aim to assess which corrections to perform in which setting and to obtain methods suitable for machine learning techniques, especially the random forest. We propose two new resampling-based methods to resemble the original data and covariance structure: stochastic inverse-probability oversampling and parametric inverse-probability bagging. We compare all techniques for the random forest and other classifiers, both theoretically and on simulated and real data. Empirical results show that the random forest profits from only the parametric inverse-probability bagging proposed by us. For other classifiers, correction is mostly advantageous, and methods perform uniformly. We discuss consequences of inappropriate distribution assumptions and reason for different behaviors between the random forest and other classifiers. In conclusion, we provide guidance for choosing correction methods when training classifiers on biased samples. For random forests, our method outperforms state-of-the-art procedures if distribution assumptions are roughly fulfilled. We provide our implementation in the R package sambia.
| S-EPMC5632994 | biostudies-literature
Project description:We studied a trend test for genetic association between disease and the number of risk alleles using case-control data. When the data are sampled from families, this trend test can be adjusted to take into account the correlations among family members in complex pedigrees. However, the test depends on the scores based on the underlying genetic model and thus it may have substantial loss of power when the model is misspecified. Since the mode of inheritance will be unknown for complex diseases, we have developed two robust trend tests for case-control studies using family data. These robust tests have relatively good power for a class of possible genetic models. The trend tests and robust trend tests were applied to a dataset of Genetic Analysis Workshop 14 from the Collaborative Study on the Genetics of Alcoholism.
| S-EPMC1866832 | biostudies-literature
Project description:Large-scale sequencing and genotyping data provide an opportunity to integrate external samples as controls to improve power of association tests. However, due to the systematic differences between genotyped samples from different studies, naively aggregating the controls could lead to inflation in Type I error rates. There has been recent effort to integrate external controls while adjusting for batch effect, such as the integrating External Controls into Association Test (iECAT) and its score-based single variant tests. Building on the original iECAT framework, we propose an iECAT-Score region-based test that increases power for rare-variant tests when integrating external controls. This method assesses the systematic batch effect between internal and external samples at each variant and constructs compound shrinkage score statistics to test for the joint genetic effect within a gene or a region, while adjusting for covariates and population stratification. Through simulation studies, we demonstrate that the proposed method controls for Type I error rates and improves power in rare-variant tests. The application of the proposed method to the association studies of age-related macular degeneration (AMD) from the International AMD Genomics Consortium and UK Biobank revealed novel rare-variant associations in gene DXO. Through the incorporation of external controls, the iECAT methods offer a powerful suite to identify disease-associated genetic variants, further shedding light on future directions to investigate roles of rare variants in human diseases.
| S-EPMC9393083 | biostudies-literature
Project description:When some individuals are screen-detected before the beginning of the study, but otherwise would have been diagnosed symptomatically during the study, this results in different case-ascertainment probabilities among screened and unscreened participants, referred to here as lead-time-biased case-ascertainment (LTBCA). In fact, this issue can arise even in risk-factor studies nested within a randomized screening trial; even though the screening intervention is randomly allocated to trial arms, there is no randomization to potential risk-factors and uptake of screening can differ by risk-factor strata. Under the assumptions that neither screening nor the risk factor affects underlying incidence and no other forms of bias operate, we simulate and compare the underlying cumulative incidence and that observed in the study due to LTBCA. The example used will be constructed from the randomized Prostate, Lung, Colorectal, and Ovarian cancer screening trial. The derived mathematical model is applied to simulating two nested studies to evaluate the potential for screening bias in observational lung cancer studies. Because of differential screening under plausible assumptions about preclinical incidence and duration, the simulations presented here show that LTBCA due to chest x-ray screening can significantly increase the estimated risk of lung cancer due to smoking by 1% and 50%. Traditional adjustment methods cannot account for this bias, as the influence screening has on observational study estimates involves events outside of the study observation window (enrollment and follow-up) that change eligibility for potential participants, thus biasing case ascertainment.
| S-EPMC5858824 | biostudies-literature
Project description:There have been recent proposals advocating the use of additive gene-environment interaction instead of the widely used multiplicative scale, as a more relevant public health measure. Using gene-environment independence enhances statistical power for testing multiplicative interaction in case-control studies. However, under departure from this assumption, substantial bias in the estimates and inflated type I error in the corresponding tests can occur. In this paper, we extend the empirical Bayes (EB) approach previously developed for multiplicative interaction, which trades off between bias and efficiency in a data-adaptive way, to the additive scale. An EB estimator of the relative excess risk due to interaction is derived, and the corresponding Wald test is proposed with a general regression setting under a retrospective likelihood framework. We study the impact of gene-environment association on the resultant test with case-control data. Our simulation studies suggest that the EB approach uses the gene-environment independence assumption in a data-adaptive way and provides a gain in power compared with the standard logistic regression analysis and better control of type I error when compared with the analysis assuming gene-environment independence. We illustrate the methods with data from the Ovarian Cancer Association Consortium.
| S-EPMC5860584 | biostudies-literature
Project description:BACKGROUND:Moderately elevated blood levels of homocysteine are weakly correlated with coronary heart disease (CHD) risk, but causality remains uncertain. When folate levels are low, the TT genotype of the common C677T polymorphism (rs1801133) of the methylene tetrahydrofolate reductase gene (MTHFR) appreciably increases homocysteine levels, so "Mendelian randomization" studies using this variant as an instrumental variable could help test causality. METHODS AND FINDINGS:Nineteen unpublished datasets were obtained (total 48,175 CHD cases and 67,961 controls) in which multiple genetic variants had been measured, including MTHFR C677T. These datasets did not include measurements of blood homocysteine, but homocysteine levels would be expected to be about 20% higher with TT than with CC genotype in the populations studied. In meta-analyses of these unpublished datasets, the case-control CHD odds ratio (OR) and 95% CI comparing TT versus CC homozygotes was 1.02 (0.98-1.07; p?=?0.28) overall, and 1.01 (0.95-1.07) in unsupplemented low-folate populations. By contrast, in a slightly updated meta-analysis of the 86 published studies (28,617 CHD cases and 41,857 controls), the OR was 1.15 (1.09-1.21), significantly discrepant (p?=?0.001) with the OR in the unpublished datasets. Within the meta-analysis of published studies, the OR was 1.12 (1.04-1.21) in the 14 larger studies (those with variance of log OR<0.05; total 13,119 cases) and 1.18 (1.09-1.28) in the 72 smaller ones (total 15,498 cases). CONCLUSIONS:The CI for the overall result from large unpublished datasets shows lifelong moderate homocysteine elevation has little or no effect on CHD. The discrepant overall result from previously published studies reflects publication bias or methodological problems.
| S-EPMC3283559 | biostudies-literature