When diseases that are suspected to be gene-influenced do not follow usual Mendelian patterns of inheritance, more than one gene, or a variety of environmental factors to which susceptible individuals are exposed, play a role. This is the case for many common diseases, including several birth defects, some forms of heart disease, hypertension, cancer, diabetes, mental illness, and infectious disorders. Evidence supporting the promise of linkage mapping for common diseases has been found in families in which the disease is inherited as a single-gene dominant disorder. For example, two different genes, each by itself responsible for Alzheimer disease have been mapped (Goate et al., 1991; Schellenberg et al., 1992). In another case, a gene thought to be responsible for only a rare form of colon cancer was later found to be implicated in the more common sporadic form of the disease (Fearon and Vogelstein, 1990). Success of the linkage approach in complex disorders has been tempered, however, by several failures. Earlier reports of linkage in manic-depressive illness in the Old-Order Amish could not be confirmed when new individuals in the family became ill (Egeland and Kidd, 1989); reports of linkage on the X chromosome previously reported for manic depression were recently retracted (Baron et al., 1992).
For disorders with complex etiologies, such as coronary artery disease, identification of a genetic defect in a rare, single-gene form may provide invaluable clues to causation and approaches to therapy in general. This has been demonstrated for familial hypercholesterolemia (Goldstein and Brown, 1989). It is possible, however, that mutations in the gene that is involved in the rare, single-gene forms of other common, complex disorders will not play a role in the complex forms. Moreover, it will be very difficult to use linkage studies to establish the role of genes that are neither necessary nor sufficient to cause the disease (Greenberg, 1993).
Prior to the discovery of recombinant DNA techniques, the determination of a person's risk of harboring genes that could lead to disease in that person, or in her or his offspring, was limited to those diseases for which a clinical diagnosis could be made or for which tests were available to detect an altered gene product (an enzyme or other protein) or a metabolite that accumulated in a readily accessible tissue such as blood, urine, skin (by biopsy), mucosa of the mouth, or hair. In the late 1960s, it proved possible to detect some altered gene products in cells obtained from the amniotic fluid by midtrimester amniocentesis, thus making prenatal diagnosis possible (Fuchs, 1966; Steele and Breg, 1966; Jacobson and Barter, 1967; Hahneman and Mohr, 1968; Nadler, 1968). However, effective treatment was not available for most prenatally diagnosable disorders. In the absence of effective treatment, prenatal diagnosis had to be done at a gestational age at which the mother had the option of aborting the affected fetus safely and legally. The number of diseases that could be diagnosed prenatally increased markedly when it became possible to localize a disease-related gene by linkage studies and