EFFECT SIZE
The extent to which a factor influences the risk of the condition under study, rather than simply an indication of whether a factor is significantly related to the condition.
by Lon R. Cardon, NATURE REVIEWS | GENETICS, VOLUME 5 | FEBRUARY 2004 | 89
The extent to which a factor influences the risk of the condition under study, rather than simply an indication of whether a factor is significantly related to the condition.
by Lon R. Cardon, NATURE REVIEWS | GENETICS, VOLUME 5 | FEBRUARY 2004 | 89
the EFFECT SIZE of a disease locus;
the frequency of the disease allele(s);
the frequency of the marker allele(s);
and the extent of LD between the marker and disease locus.
These four parameters are the result of the more subtle hallmarks of MULTIFACTORIAL DISEASES, including interactions among disease loci that are related to effect size or the fact that the disease allele frequencies might reflect a range of mutations at the same locus. This set of parameters provides a convenient summary of the basic aim of the association-study design: to correlate genotypes and disease phenotypes that are obtained from a sample of individuals. In this review, we discuss how we can optimize our chances of finding complex disease associations by examining the interplay of the factors that influence the size of an observed association and, therefore, our ability to find associations. We focus here on genetic variants that are amenable to detection in population-based association studies; that is, we exclude those that are so rare (<0.01–0.001 in frequency) that only family-based studies would realistically provide sufficient numbers of cases to explore the association.
The case–control study and its effect size
The standard measure of effect in the case–control study is the odds ratio (OR), defined as the odds of exposure among cases divided by the odds of exposure among controls.
The OR provides a good approximation of the relative risk of the risk factor in question (that is, the ratio of risk of disease in people with the risk factor to that in people without) if sampling for the case–control study mimics that of a prospective cohort study.
