Gene and Environment

As the example of PKU demonstrates, studies that attempt to identify factors important in determining human phenotypes must simultaneously examine multiple genetic and environmental factors. It is generally not possible to experiment directly on humans to observe the effects of a gene or environmental factor on the expression of a phenotype. Human genetic studies are generally observational studies, where the researcher is limited to observing the combinations of exposures that naturally occur in the population. Genetic epidemiologists must apply statistical methods to these observational data to evaluate how genes and the environment affect the development of a phenotype.

The simultaneous analysis of genetic and environmental factors allows the identification of environmental factors that interact with each other. Researchers use statistical models to compare the joint effects of genetic and environmental factors in people with the phenotype and people without the phenotype. Such "case-control studies" are commonly used to examine the relationship between disease phenotypes and both genetic and nongenetic risk factors.

The strength of the association between a risk factor and the disease is described by the "relative risk," which is the probability of having the phenotype if exposed to the risk factor divided by the probability of having the phenotype if not exposed to the risk factor. A relative risk greater than 1 suggests that the risk factor increases the probability of developing the phenotype, whereas a relative risk less than 1 suggests the risk factor decreases the probability of the phenotype. An association between a risk factor and a phenotype exists if the relative risk is significantly different from 1.

An example of a risk factor and relative risk can be seen in Alzheimer's disease (AD). One gene that influences the risk of AD is the APOE gene. Three common alleles are known: ε-2, ε-3, and ε-4. Caucasian Americans with one or more copies of ε-4 are two and one-half times more likely to develop AD than are people with two ε-3 alleles. Interestingly, the ε-4 allele is not as strong of a risk factor for AD among African Americans or Hispanics, who nonetheless have higher risks of developing AD than Caucasians.