Limits...
The importance of distinguishing between the odds ratio and the incidence rate ratio in GWAS.

Waltoft BL, Pedersen CB, Nyegaard M, Hobolth A - BMC Med. Genet. (2015)

Bottom Line: In recent years, genome wide association studies have identified many genetic variants that are consistently associated with common complex diseases, but the amount of heritability explained by these risk alleles is still low.We find significant numerical differences between the odds ratio and the incidence rate ratio when the fact that gene variant may be associated with competing events, e.g. lifetime, is ignored.The ranking of the SNPs according to p-values may differ between the two study designs.

View Article: PubMed Central - PubMed

Affiliation: National Center for Register-based Research, Department of Economics and Business Economics, Aarhus University, Fuglesangs allé 4 room K10, 8210, Aarhus V, Denmark. berit@econ.au.dk.

ABSTRACT

Background: In recent years, genome wide association studies have identified many genetic variants that are consistently associated with common complex diseases, but the amount of heritability explained by these risk alleles is still low. Part of the missing heritability may be due to genetic heterogeneity and small sample sizes, but non-optimal study designs in many genome wide association studies may also have contributed to the failure of identifying gene variants causing a predisposition to disease. The normally used odds ratio from a classical case-control study measures the association between genotype and being diseased. In comparison, under incidence density sampling, the incidence rate ratio measures the association between genotype and becoming diseased. We estimate the differences between the odds ratio and the incidence rate ratio under the presence of events precluding the disease of interest. Such events may arise due to pleiotropy and are known as competing events. In addition, we investigate how these differences impact the association test.

Methods: We simulate life spans of individuals whose gene variants are subject to competing events. To estimate the association between genotype and disease, we applied classical case-control studies and incidence density sampling.

Results: We find significant numerical differences between the odds ratio and the incidence rate ratio when the fact that gene variant may be associated with competing events, e.g. lifetime, is ignored. The only scenario showing little or no difference is an association with a rare disease and no other present associations. Furthermore, we find that p-values for association tests differed between the two study designs.

Conclusions: If the interest is on the aetiology of the disease, a design based on incidence density sampling provides the preferred interpretation of the estimate. Under a classical case-control design and in the presence of competing events, the change in p-values in the association test may lead to false positive findings and, more importantly, false negative findings. The ranking of the SNPs according to p-values may differ between the two study designs.

Show MeSH

Related in: MedlinePlus

The density of the three outcomes. The densities of the three types of outcome; death and two different diseases. The black solid line in subfigure A is the density function of a Gompertz distribution with density mode at age 85 and a shape parameter of 0.0004, i.e. the overall baseline. This is the likelihood of dying given an individual’s age. The black solid line in subfigures for Disease 1 and Disease 2 is the density function of a Gompertz distribution with density mode at age 25 and a shape parameter of 0.95 and mode at age 55 and shape parameter 0.1 respectively. This is the likelihood of getting diseased given an individual’s age. The coloured lines correspond to the likelihood of dying or developing the disease for different genotypes and different associations between the SNP and death
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4593225&req=5

Fig2: The density of the three outcomes. The densities of the three types of outcome; death and two different diseases. The black solid line in subfigure A is the density function of a Gompertz distribution with density mode at age 85 and a shape parameter of 0.0004, i.e. the overall baseline. This is the likelihood of dying given an individual’s age. The black solid line in subfigures for Disease 1 and Disease 2 is the density function of a Gompertz distribution with density mode at age 25 and a shape parameter of 0.95 and mode at age 55 and shape parameter 0.1 respectively. This is the likelihood of getting diseased given an individual’s age. The coloured lines correspond to the likelihood of dying or developing the disease for different genotypes and different associations between the SNP and death

Mentions: We use a Cox proportional hazard model where the baseline hazards follow a Gompertz distribution [14] for both events. The baseline for the incidence of dying follows a Gompertz distribution with mode 85 and shape parameter 0.0004. The densities for different effects of the number of minor alleles on the IRR of death are shown in Fig. 2. Two different parameter settings are used for the baseline of disease: one with mode 25 and shape parameter 0.95, and one with mode 50 and shape parameter 0.1. Figure 2 shows the densities for the two parameter settings for different effects of the number of minor alleles on the IRR. The black lines in Fig. 2 correspond to the overall baseline density. We multiply the incidence rate of disease by a constant (less than 1) in order to scale the life-time risk of disease [15]. Different values of the constant are used to consider common versus rare diseases; the smaller the constant the rarer the disease. We assume a log linear proportional effect of the number of minor alleles on the incidence rate of disease and death, with different incidence rate ratios of 0.5, 1.0, 1.1, 1.2, 1.5, 1.7, 2.0 and 3.0 for disease and 0.5, 0.75, 0.9, 1.0, 1.1, 1.2, 1.5, 1.7, 2.0, and 3.0 for death.Fig. 2


The importance of distinguishing between the odds ratio and the incidence rate ratio in GWAS.

Waltoft BL, Pedersen CB, Nyegaard M, Hobolth A - BMC Med. Genet. (2015)

The density of the three outcomes. The densities of the three types of outcome; death and two different diseases. The black solid line in subfigure A is the density function of a Gompertz distribution with density mode at age 85 and a shape parameter of 0.0004, i.e. the overall baseline. This is the likelihood of dying given an individual’s age. The black solid line in subfigures for Disease 1 and Disease 2 is the density function of a Gompertz distribution with density mode at age 25 and a shape parameter of 0.95 and mode at age 55 and shape parameter 0.1 respectively. This is the likelihood of getting diseased given an individual’s age. The coloured lines correspond to the likelihood of dying or developing the disease for different genotypes and different associations between the SNP and death
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4593225&req=5

Fig2: The density of the three outcomes. The densities of the three types of outcome; death and two different diseases. The black solid line in subfigure A is the density function of a Gompertz distribution with density mode at age 85 and a shape parameter of 0.0004, i.e. the overall baseline. This is the likelihood of dying given an individual’s age. The black solid line in subfigures for Disease 1 and Disease 2 is the density function of a Gompertz distribution with density mode at age 25 and a shape parameter of 0.95 and mode at age 55 and shape parameter 0.1 respectively. This is the likelihood of getting diseased given an individual’s age. The coloured lines correspond to the likelihood of dying or developing the disease for different genotypes and different associations between the SNP and death
Mentions: We use a Cox proportional hazard model where the baseline hazards follow a Gompertz distribution [14] for both events. The baseline for the incidence of dying follows a Gompertz distribution with mode 85 and shape parameter 0.0004. The densities for different effects of the number of minor alleles on the IRR of death are shown in Fig. 2. Two different parameter settings are used for the baseline of disease: one with mode 25 and shape parameter 0.95, and one with mode 50 and shape parameter 0.1. Figure 2 shows the densities for the two parameter settings for different effects of the number of minor alleles on the IRR. The black lines in Fig. 2 correspond to the overall baseline density. We multiply the incidence rate of disease by a constant (less than 1) in order to scale the life-time risk of disease [15]. Different values of the constant are used to consider common versus rare diseases; the smaller the constant the rarer the disease. We assume a log linear proportional effect of the number of minor alleles on the incidence rate of disease and death, with different incidence rate ratios of 0.5, 1.0, 1.1, 1.2, 1.5, 1.7, 2.0 and 3.0 for disease and 0.5, 0.75, 0.9, 1.0, 1.1, 1.2, 1.5, 1.7, 2.0, and 3.0 for death.Fig. 2

Bottom Line: In recent years, genome wide association studies have identified many genetic variants that are consistently associated with common complex diseases, but the amount of heritability explained by these risk alleles is still low.We find significant numerical differences between the odds ratio and the incidence rate ratio when the fact that gene variant may be associated with competing events, e.g. lifetime, is ignored.The ranking of the SNPs according to p-values may differ between the two study designs.

View Article: PubMed Central - PubMed

Affiliation: National Center for Register-based Research, Department of Economics and Business Economics, Aarhus University, Fuglesangs allé 4 room K10, 8210, Aarhus V, Denmark. berit@econ.au.dk.

ABSTRACT

Background: In recent years, genome wide association studies have identified many genetic variants that are consistently associated with common complex diseases, but the amount of heritability explained by these risk alleles is still low. Part of the missing heritability may be due to genetic heterogeneity and small sample sizes, but non-optimal study designs in many genome wide association studies may also have contributed to the failure of identifying gene variants causing a predisposition to disease. The normally used odds ratio from a classical case-control study measures the association between genotype and being diseased. In comparison, under incidence density sampling, the incidence rate ratio measures the association between genotype and becoming diseased. We estimate the differences between the odds ratio and the incidence rate ratio under the presence of events precluding the disease of interest. Such events may arise due to pleiotropy and are known as competing events. In addition, we investigate how these differences impact the association test.

Methods: We simulate life spans of individuals whose gene variants are subject to competing events. To estimate the association between genotype and disease, we applied classical case-control studies and incidence density sampling.

Results: We find significant numerical differences between the odds ratio and the incidence rate ratio when the fact that gene variant may be associated with competing events, e.g. lifetime, is ignored. The only scenario showing little or no difference is an association with a rare disease and no other present associations. Furthermore, we find that p-values for association tests differed between the two study designs.

Conclusions: If the interest is on the aetiology of the disease, a design based on incidence density sampling provides the preferred interpretation of the estimate. Under a classical case-control design and in the presence of competing events, the change in p-values in the association test may lead to false positive findings and, more importantly, false negative findings. The ranking of the SNPs according to p-values may differ between the two study designs.

Show MeSH
Related in: MedlinePlus