Human Genomics Research: New Challenges for Research Ethics

The 1975 Asilomar Conference dealt with two different kinds of threat: a threat to human safety and, implicitly, a threat to the independence of scientists. The moratorium it called on some kinds of research with recombinant DNA made it possible for the safety of such research to be more thoroughly investigated. It also made it possible for biologists to present themselves as worthy of public trust. Both results were important to the future of research in molecular biology.

Today, a branch of science whose tools descend, in large part, from that recombinant DNA research faces a similar double threat. Research into human genetics has stretched current regulations of human subjects research beyond the breaking point. In the context of this kind of research, those regulations, while they largely protect the safety of human subjects, no longer protect their interests. And, as a result, research conducted under those regulations risks violating the public's trust and thus imperiling future studies in human genetics.

Problems in human subjects protections arise in two different contexts. One set of special problems concerns research conducted with "groups" of people, groups that have a pre-existing cultural significance. The second set concerns a broader type of research that seeks associations between genetic variations and human health in large populations through the development of vast databases of phenotypic and genotypic information. Each type of research, as currently regulated, risks leaving its human subjects feeling cheated and embittered.

Research with Culturally-Defined Groups

The issue of "group" or "community" concerns in genetics research stems from one common approach to human genetics research. Researchers identify a somewhat discrete human population that has a higher than average incidence of a disease or condition thought to be linked to genetic variation. The higher incidence in the population might make the genetic links easier to find--as may the likely greater similarities in the group's environment, broadly defined. Thus are born studies of schizophrenia among the Old Order Amish, non-insulin dependent diabetes mellitus among the Pima, or asthma on Tristan da Cuhna. Similarly, those interested in using patterns of genetic variation as evidence of the history of populations will examine genetic markers from individuals from different ethnic or cultural groups. These two kinds of group-based research lead to publications about "diabetes among the Pima" or the "ancestry of the Han." And therein lies the problem.

Genetics research on groups that are, or are perceived as being, closely genetically connected has implications for all members of those groups, whether or not they decided--or even were asked--to take part in the research. Sometimes, those implications might be positive, such as when a good medical intervention is found to aid the health of the group. Other times, the results may be negative, leading to stigmatization of or discrimination against the group involved--or to a change in the group's culture through contradiction of the group's own historical knowledge. Most of the time, the research will have little effect, one way or the other. But, whatever the effects, they will land on many group members who did not give their informed consent after weighing the foreseeable benefits and risks.

This issue was first confronted directly in the context of the Human Genome Diversity Project. The North American Regional Committee of that Project opted in its Model Ethical Protocol for the Collection of DNA to require, when feasible, the consent of the "group" to such group research (North American Regional Committee 1997; Greely 1997a; Greely 1997b). The resulting discussion of how to protect the interests of groups or communities in research has now spawned a substantial and growing literature (Weijer et al. 1999), some of it in favor of some kind of group consultation or consent (Foster et al. 1997; Foster et al. 1998; Foster et al. 1999; Sharp 2000; Freeman, 1998) and some of it in opposition (National Research Council 1997; Juengst 1998a...