These chapters were published with modifications by Oxford University Press, 2004

Ethical, Legal, and Social Issues in the Design and Conduct of Human Genome Epidemiologic Studies
Laura M. Beskow, MPH, Department of Health Policy and Administration, University of North Carolina at Chapel Hill School of Public Health

| References |

Introduction
Human genetic variation provides an important tool for refining epidemiologic research. Although investigators have long recognized that disease generally results from a constellation of host- and environment-specific factors, they have tended to focus on the environment due to scientific and technologic limitations. Environmental exposures alone, however, whether chemical, physical, infectious, nutritional, social, or behavioral, often poorly predict who will develop disease--particularly the common, complex diseases of public health interest. Understanding the impact of environmental factors among people who have specific genetic susceptibilities offers the possibility of more effective and targeted medical and public health interventions. A critical question is: How can we advance this public health research agenda while protecting and respecting the participants in such research?

Ethical principles generally require that epidemiologists working with human participants submit their research protocols for independent review, obtain informed consent, protect privacy and maintain confidentiality, and safeguard the rights and welfare of the individuals and groups they study [1]. How these obligations are best addressed depends on the nature of the risks and benefits associated with the research. Whether and how the addition of a genetic component to an epidemiologic study changes the balance of risks and benefits merits further analysis. Many argue that genetic information is fundamentally similar to other kinds of health information [2]. Even so, society invests enormous power in the concept of genetics, and notions of genetic determinism and genetic reductionism have significant negative implications for public health and prevention messages [3]. Combined with the history of eugenics and other research abuses in the U.S. and around the world, clarifying the duties of investigators to participants in population-based research involving genetics is important. Although much has been written about ethical, legal, and social issues associated with genetic testing [4,5], and about ethical issues in epidemiology [6,7], less attention has been given to issues that arise specifically in the context of genetic epidemiology.

This chapter explores ethical, legal, and social issues in human genome epidemiologic research, highlighting differences in the risks and benefits of genetic research among high-risk families compared to population-based settings. This exploration draws on guidelines for research involving human tissue, such material being necessary for conducting genetic analyses in any context. (See also 8 for an international perspective).

Background: Protection of Human Participants in Research
Regulations and institutional rules for the protection of human research participants derive from a number of sources. Among the best known of these is the Belmont Report [9], which identifies three ethical principles particularly relevant to research involving human participants: respect for persons, beneficence, and justice. Federal policy for the protection of human research participants incorporates these principles in two basic protections: oversight by an Institutional Review Board (IRB) and requirements for informed consent. These policies, codified by the Department of Health and Human Services in the Code of Federal Regulations, Title 45, Part 46, Subpart A (45 CFR 46), are often called the “Common Rule” because an identical set of regulations has been adopted by a number of federal agencies. A diverse array of research is subject to the Common Rule, involving varying combinations of risks and benefits. The regulations therefore provide for exemption, waiver, or alteration of the requirements for review and informed consent when appropriate.

Family-Based Studies vs. Population-Based Studies Involving Genetics
Genetic research is typically considered a sensitive enterprise because much of it has been conducted among families with a heavy burden of disease, e.g., the investigation of BRCA1 mutations among families with multiple members affected with breast or ovarian cancer. Family studies provide a unique framework for investigating highly penetrant gene variants, which are those that lead to disease expression a majority of the time and thus produce marked familial aggregation. However, this kind of research poses the risk of psychological and social harms, such as insurance and employment discrimination, social stigmatization, familial disruption, and psychological distress, which could result from uncovering genetic information that has significant implications for the health of the individual and her family. These concerns are magnified when only limited or unproven measures are available for prevention or treatment. Thus, recommendations for the protection of research participants developed in this context usually call for close IRB oversight, detailed informed consent disclosures, and counseling by a genetic counselor or medical geneticist [10].

Highly penetrant gene variants ideally should also be studied in a population-based setting. Although BRCA1 mutations are thought to account for less than 5% of all breast cancer [11], quantifying their impact in the general population could help scientists understand the risks, mechanisms, and natural history of breast cancer in general. The most important contribution of population-based research, however, will be to elucidate the interactions between common, lower-penetrance gene variants and modifiable environmental factors that influence the probability of disease. Identifying these kinds of genetic factors may help illuminate the biologic processes responsible for disease, leading to new avenues for prevention or treatment. Increasing knowledge about gene-environment interactions may one day enable us to provide individuals with more accurate predictions of their disease risk as well as more effective and tailored interventions--such as more frequent or earlier medical surveillance, lifestyle or dietary modifications, or targeted drug therapy [12].

To move this research agenda forward, it is important to have guidelines to protect participants from risks specifically associated with integrating genetic variation into population-based research. Such research would be expected to have meaningful public health implications but involve a low probability of physical, psychological, or social harm to individual research participants. Guidelines developed in the context of family-based research generally fail to distinguish between research on gene variants known to carry high risk for disease and research on lower-risk genotypes that would allow a better understanding of disease processes and the role of environmental exposures [13]. As noted by Clayton et al. [14], the risks involved in identifying high-penetrance mutations must be distinguished from the risks of identifying “common alleles that are neither necessary nor sufficient for the development of disease.” The uniform application of guidelines for research on high-risk genotypes to all research with a genetic component, without adjustment for the type or degree of risk involved, could render many studies of gene-environment interaction impossible to conduct and deprive society of important and beneficial knowledge.

Application of Guidelines for Research Involving Human Tissue
One approach to clarifying the obligations of investigators to participants in population-based research involving genetics is to draw on guidelines for research involving human tissue. Genetic analysis in any context requires human biological material. Numerous collections of such material already exist; public health examples include residual dried bloodspots from newborn screening [15] and specimens collected as part of population-based studies, such as the National Health and Nutrition Examination Survey (NHANES) [16] and the Framingham [17] and Jackson [18] heart studies. New public health collections will be created at an increasing rate as their value for genetic research is recognized; for example, the storage of specimens collected during the investigation of infectious diseases [19].

Many groups have published statements regarding the appropriate research use of human tissue [12,20,21,22]. The National Bioethics Advisory Commission (NBAC) examined this topic in its August 1999 report, Research Involving Human Biological Materials: Ethical Issues and Policy Guidance [23]. This report interpreted several concepts and terms in the Common Rule and recommended ways to both strengthen and clarify existing regulations with regard to research using human tissue.

The following discussion does not provide a comprehensive review of the extant literature on human tissue research, but uses NBAC’s recommendations to highlight some of the ethical, legal, and social issues that arise specifically in the context of research on the interactions between lower-penetrance gene variants and environmental factors.

ISSUES IN HUMAN GENOME EPIDEMIOLOGIC RESEARCH
Is the Research Subject to Human Participants Regulations?
For investigators conducting research using human biological materials, knowing whether and how policies intended to protect human participants apply may sometimes be difficult. With certain exceptions described below, the Common Rule applies to virtually all research involving human participants conducted or supported by the federal government (45 CFR 46.101(a)). Non-federally funded research may be controlled by federal policy if the research is subject to regulation by the Food and Drug Administration, or if it is conducted at an institution that has voluntarily agreed to apply Common Rule requirements to all its research.

“Human participant” is defined as a living individual about whom an investigator conducting research obtains (1) data through intervention or interaction with the individual, or (2) identifiable private information (45 CFR 46.102(f)). Population-based research in which investigators collect new biological samples for genetic analysis clearly entails an interaction with individual participants. Population-based research using existing biological samples may not require an intervention or interaction but may involve identifiable information to correlate genetic findings with health outcomes. For instance, investigators used samples and health data from the Arteriosclerosis Risk in Communities cohort to measure the associations between apolipoprotein E polymorphisms, physical and lifestyle factors, and the prevalence of significant carotid artery Arteriosclerosis [24]. Still other kinds of population-based research can be conducted using existing biological materials with no identifiable information and no interaction with participants. One example is a study in New York for which researchers used anonymized residual newborn blood spots to determine in different populations the prevalence of two interleukin-4 receptor polymorphisms thought to be associated with asthma [25]. In all of these studies, the risks to participants are primarily related to the disclosure of private information, and the level of protection needed is directly related to the ease and likelihood of identifying the sources of the samples. NBAC’s report thus recommends that the Common Rule be interpreted such that the following categories of biological samples may be used without informed consent or IRB review:

Samples that are unidentified. According to NBAC, unidentified (or “anonymous”) samples are those created from specimens for which identifiable personal information was never collected or was not maintained and cannot be retrieved by the repository. They advise that research using unidentified samples does not meet the definition of human participants research and is not regulated by the Common Rule (NBAC Rec. 1A).

Samples that have been unlinked through a sound process. Research conducted with unlinked (or “anonymized”) samples is human participants research and is regulated by the Common Rule but is eligible for exemption from IRB review pursuant to 45 CFR 46.101(b)(4) (NBAC Rec. 1B). According to this section of the federal regulations, research activities that are eligible for exemption include:

Research involving the collection or study of existing data, documents, records, pathological specimens, or diagnostic specimens, if these sources are publicly available or if the information is recorded by the investigator in such a manner that subjects cannot be identified, directly or through identifiers linked to the subjects [emphasis added].

See “Unlinking Samples” for further discussion of risks and potential benefits associated with anonymizing samples.

Samples that are publicly available. Research using coded or directly identified samples is human participants research and, because participants can be identified, it is not eligible for the above-referenced exemption unless the specimens are publicly available (NBAC Rec. 1C). Generally, the justification for exempting research using publicly available data or specimens is that investigators are not obtaining private information when examining resources that are already in the public domain, and any harm from its disclosure has already occurred. However, exactly what constitutes “publicly available” under the Common Rule has been a matter of some confusion. For example, some may consider NHANES specimens to be a public resource because they are collected and banked by an agency of the federal government. However, samples from NHANES are in fact available only to qualified researchers and federal laws strictly protect the confidentiality of all NHANES health data and samples [16]. NBAC concludes that in the case of human biological materials, the term “publicly available” should be defined literally as only those materials available to the general public [p. 59]. They also note that any data that emerges from genetic analysis of stored tissue was not previously “existing” in any genuine sense, much less publicly available. Thus, the exemption for research using existing specimens from a publicly available source will rarely, if ever, apply to population-based research involving genetics.

Samples from participants who are deceased. Because the Common Rule defines a human participant as a living person, there is no human participant if the source of the biological material is deceased. No IRB review or informed consent is required for research using samples from participants who are deceased, but NBAC suggests that because research on such samples may have implications for living relatives, investigators should to the extent possible plan their studies to avoid harms to these individuals (NBAC Rec. 17).

Although NBAC notes legal possibilities of exemption under the Common Rule, many institutions require some form of IRB review even for technically exempt studies. The goal of IRB oversight is to ensure that risks to participants are minimized, that risks are reasonable in relation to anticipated benefits, that selection of participants is equitable, and that informed consent procedures are adequate (45 CFR 46.111(a)). It can be problematic to charge researchers with deciding that their own projects do not need IRB review; in some cases, the potential for conflict of interest or lack of detailed knowledge about federal regulations or local policies may exist [26]. The Common Rule provides for an expedited review procedure, under which the IRB chairperson or designee can exercise all of the authorities of the IRB except disapproving the research (45 CFR 46.110).

Unlinking Samples
Because the risks of genetic research are related to the disclosure of private information, unlinking biological samples from identifying information essentially removes the potential for harm. Truly “anonymizing” genetic information may not be possible because one could theoretically always match a known person to his or her research sample. However, at present, it would be extremely difficult to match a research sample back to a particular person without identifiers or another DNA sample from that person.

However, unlinking samples also raises a number of concerns. These include the administrative cost of making the samples completely unidentifiable [27], the possibility of compromising the value of the research [28], and the potential ethical objection that investigators have the opportunity to seek consent but instead unlinked the samples [14]. These concerns may be especially relevant for population-based research involving genetics. For example, the value of some studies may be reduced if the only way to render the samples unidentifiable is to limit linkage to demographic and exposure data. An investigation of an outbreak of leptospirosis among tri-athletes [28,29] offered a unique situation in which a number of exceedingly healthy people were nearly uniformly exposed to contaminated water, thereby facilitating research on genetic susceptibility factors. However, consent for such research had not been obtained and all information regarding previously identified environmental risk factors (swim time and approximate amount of water ingested) would have had to be dropped to completely unlink the samples.

According to NBAC, when a researcher proposes to create unlinked samples from identifiable materials already under his or her control, the research may be exempt from IRB review if (1) the process used to unlink the samples will be effective, and (2) unlinking the samples will not unnecessarily reduce the value of the research (NBAC Rec. 3). NBAC did not put forward a preferred method for unlinking samples but instead chose the word “effective” to allow for IRB judgment in ensuring the appropriateness of any method to reduce the risk of harm. There is a continuum between directly identified and completely anonymous tissues or data, many points on which can deter the ability to identify participants whose tissue or data remain in some sense linked [21] (see “Privacy and Confidentiality” below). One should not assume that risk can or must be completely eliminated. As Buchanan [27] notes, “It is a mistake to proceed on the assumption that the goal is to develop policies that reduce the risk…to zero, as if risk reduction were costless,” i.e., as if the costs of reducing risk did not include setbacks to important and morally legitimate interests, including improving health and preventing harm through the application of scientific knowledge in health care.

Assessing Risk
The Belmont Report formulated two complementary expressions of the ethical principle of beneficence: (1) do not harm, and (2) maximize possible benefits and minimize possible harms [9]. A major responsibility of IRBs is to assess the risks and anticipated benefits of proposed research, both to participants and to society [32], and a part of this assessment is to determine whether the research presents greater than minimal risk. Minimal risk is a pivotal concept with regard to eligibility for expedited review and to the possibility of waiver or alteration of consent requirements.

In addition, many population-based studies will focus on lower-penetrance gene variants. Lower-penetrance gene variants by definition lead to smaller increases in relative risk for disease and, absent misplaced notions of genetic determinism, a corresponding decrease in the probability of harms stemming from the misuse of the information. Steinberg et al. [34] note that testing identifiable specimens for relatively benign polymorphisms that have a small impact on disease risk should entail less risk for loss of insurance, psychological distress, or social stigmatization than testing for the genetic mutations that almost ensure development of a serious disease. Others similarly argue that risk factors such as blood pressure or cholesterol level show comparable patterns of incomplete penetrance and there is little reason that risk factors based on DNA should not be treated the same way [35]. Thus, when assessing risk, it is important not only to assess the likelihood that individual results could be inappropriately disclosed, but also the probability and magnitude of the harm that could realistically result if such disclosure occurred. NBAC suggests that IRBs consider the following questions when determining the extent to which the source of a sample could be harmed [p. 67]:

• How easily identifiable is the source?
• What is the likelihood that the source will be traced?
• If the source is traced, what is the likelihood that persons other than the investigators will obtain information about the source?
• If non-investigators obtain information regarding the source, what is the likelihood that harm will result?

NBAC recommends that IRBs operate on the presumption that research on coded samples is of minimal risk when adequate safeguards are in place to protect confidentiality (NBAC Rec. 10).

Privacy and Confidentiality
According to the Office for Human Research Protections’ IRB Guidebook [32], privacy can be defined as having control over the extent, timing, and circumstances of sharing oneself (physically, behaviorally, or intellectually) with others. Confidentiality pertains to the treatment of information that an individual has disclosed in a relationship of trust with the expectation that it will not be divulged to others in ways that are inconsistent with the understanding of the original disclosure without permission.
Risks that arise from the potential for invasion of privacy or breach of confidentiality may pose the possibility of serious harm, as could happen if sensitive personal information is disclosed to third parties who misuse the information to discriminate against or stigmatize research participants. These kinds of concerns are often expressed with regard to genetic information, but they are also of concern with regard to many other kinds of health information, e.g., a diagnosis of angina or diabetes, HIV status, history of treatment for drug addiction. Informational risks may also take the form of “wrongs,” rather than harms, such as treating people solely as a means to an end [36]. Wrongs are often perceived to be less grave than harms, although they may infringe on the ethical principles of respect for persons and beneficence. When addressing harms and wrongs, it is important to try to craft a balance that respects individual interests in privacy and avoiding discrimination and societal interests in clinical benefit, research, and public health [2].

Assessing the adequacy of safeguards to protect the privacy and confidentiality of all health information, including genetic information, requires attention to a number of factors. These include physical, technological, organizational, and administrative practices and procedures, as well as legal protections (e.g., Certificates of Confidentiality). Detailed manuals are available from agencies such as the National Center for Health Statistics of the Centers for Disease Control and Prevention that have extensive experience in collecting and maintaining confidential health information [37]. Absolute privacy of health information can never be ensured, even with maximal security protections, because no system can safeguard against access by those who are authorized to use the data system [2]. Consent documents should provide details about measures that will be taken to protect privacy, as well as disclose that confidentiality cannot be guaranteed.

Informed Consent
The process of developing and implementing informed consent procedures is intended to ensure that each prospective participant receives sufficient information, in a comprehensible manner and under conditions free of coercion and undue influence, to make an autonomous choice to participate or not participate in a research project. It also reminds researchers of their ethical obligations to participants [45]. Once research is deemed to involve human participants, informed consent is required unless all four of the following criteria are met (45 CFR 46.116(d)):

The research involves no more than minimal risk to the participants. See “Assessing Risk” above.
The waiver or alteration will not adversely affect the rights and welfare of the participants. Given the difficulties of interpreting minimal risk, the application of this criterion is vital in determining the appropriateness of a waiver [46]. NBAC elaborates a number of ways in which the rights of research participants could be compromised [p. 68]. For instance, participants’ interests in controlling sensitive information and not being discriminated against could be jeopardized in some studies if existing samples are used without consent. Both participants’ and non-participants’ interests may be harmed by research that explicitly compares, for example, ethnic or racial groups; see “Group Harms” above. These concerns do not mean such studies should not be undertaken, but that they should be considered carefully before a waiver of consent is granted (NBAC Rec. 11).

The research could not practicably be carried out without the waiver or alteration. As with “publicly available,” defining “practicably” can be difficult. Depending on the prevalence of the gene variant, the frequency of exposure to environmental risk factors, and the magnitude of the effect to be detected, gene-environment interaction studies may require hundreds or thousands of biological samples [47]. Existing stores of specimens represent a potentially valuable and efficient resource. Requiring without exception that the sources of these samples be recontacted to gain informed consent for each use could delay or halt some important research. At the same time, obtaining informed consent is a reflection of the ethical principle of respect for persons and the conviction that individuals who are capable of self-determination should be treated as autonomous agents [9]. NBAC believes that the criteria of minimal risk and rights and welfare should be the compelling considerations in determining the appropriateness of waiving consent [p. 69]. They suggest (NBAC Rec. 12) that IRBs may presume that meeting consent requirements would be impracticable when using existing coded or identified materials and when the research poses minimal risk. For new materials collected after the adoption of NBAC’s recommendations, IRBs should apply the recommended informed consent process and their usual standards for the practicability requirement.

Some disagree with the portion of this recommendation dealing with existing materials because of the challenges in evaluating minimal risk, and because it forfeits the opportunity for feedback from research participants in cases in which it is practicable to obtain consent [46]. Grizzle et al. [22], for example, agree that IRBs should be permitted broader latitude to waive the requirement for informed consent--provided that written nondisclosure, confidentiality, and security policies have been approved--but that such a waiver should be granted on a case-by-case basis.

Whenever appropriate, the subjects will be provided with additional pertinent information after participation. NBAC finds that this debriefing requirement, which has a historical basis in “deception studies” used in behavioral sciences, is usually not relevant to research involving human biological materials and may even cause harm (NBAC Rec. 13).

Consent Disclosures
The Common Rule specifies eight required elements of information (and six optional elements) that must be conveyed to prospective participants (45 CFR 46.116). Because genetic research evolved in the context of family studies of highly penetrant genetic mutations, recommendations concerning specific disclosures have understandably focused on the consequences of discoveries that have implications for the health of individual participants and their families. For example, the American Society of Human Genetics in 1996 recommended that disclosures to individuals considering participation in genetic studies should include, among other things, the potential implications of research results, the possibility of unexpected findings such as medical risks, carrier status, risks to offspring, and misidentified parentage, and the possibility of adverse psychological sequelae, disruption of family dynamics, social stigmatization and discrimination [10].

In studies of diseases involving low-penetrance gene variants and environmental exposures, where the possibility of confirmed and clinically relevant findings are remote, requiring some of these disclosures may be inappropriate. The investigator’s charge is to neither understate nor overstate the risks involved [48], and tailoring the content of informed consent disclosures to convey risks and benefits that are material to the research at hand will promote prospective participants’ ability to make informed choices about entering the study.

Consent for Storage and Future Research
When human biological materials are collected, whether in a research or clinical setting, asking participants for their consent is appropriate if secondary uses are foreseen [23,32]. Consent to future research is meaningful, however, only if participants appreciate the types of studies that may be conducted [23]. NBAC recommends that consent forms be developed that provide a sufficient number of options to help prospective participants understand clearly the decision they are asked to make and presents six possible options (NBAC Rec. 9):

• refusing use of their biological materials in research;
• permitting unidentified or unlinked use only;
• permitting coded/identified use for one study only;
• permitting coded/identified use for one study and contact regarding further studies;
• permitting coded/identified use for any studies related to the condition for which the sample was originally collected; or
• permitting coded/identified use for any future study.

In some situations, offering this number of relatively imprecise options may be prohibitively complex. One alternative is for consent documents to state that investigators would like to store remaining samples for future research, describe plans for such research to the extent they are known, and provide participants the opportunity to consent or refuse [44]. The description of future research plans should include whether the samples will be unlinked, coded, or directly identified. It could also include whether studies using the samples will be confined to a particular disease or class of diseases, who (or at least what types of investigators) will have access to the samples, and whether these investigators would be given unlinked or coded samples. Storing samples for future research is essentially a separate project, and thus consent forms should expressly state the right to refuse to have one’s sample stored irrespective of the decision to participate in the immediate project [48].

The matter of consent for any future study will no doubt continue to engender controversy. One NBAC commissioner voiced concerns regarding the last two options set forth in Recommendation 9 because he believes neither adequate IRB review nor informed consent is possible in the context of unspecified future studies entailing unknown risks and benefits [p. 65]. Public input about the individual and societal risks and benefits of population-based research involving genetics would be a valuable contribution to this debate. Little information is available about the opinions of the public at large or of minority groups about research uses of human tissue [49], and specifically about the storage and use of biological samples for gene-environment interaction research and the application of genetics in disease prevention. In addition, as Feinleib [50] noted with regard to how explicit the description of proposed research must be to qualify for fully informed consent: “This is an eminently researchable area in its own right and would give valuable insight into how far it is desirable to go on providing full disclosures, what aspects are most persuasive to those who consent to participate, and what are the concerns of those who decline.”

Reporting Research Results
Whether and when to disclose individual research results to participants are additional areas of debate. Although it is essentially universally accepted that individuals should have access to their own medical information, several key characteristics differentiate medical information from research data [51]. However, some IRBs have held that investigators are obligated to offer participants in genetic research their individual results. This stance is perhaps based in part on justifiable concerns arising in the context of family-based research, but it can create serious problems when applied to some population-based studies.

First, the clinical validity of the results will not be known until a chain of evidence regarding risk associations has been established. Grizzle et al. [22] state, “A single research project does not establish irrefutable scientific fact, and the results of a single investigation have no applicability to an individual patient. Disclosure of a single research project’s results to a patient is at best not beneficial, and at worst could be misleading or even harmful.” Similarly, Armstrong [52] notes, “By definition, the study of low-penetrance genes seeks to find small effects that may be observed only in conjunction with certain exposures or in certain population subgroups. Dismissing results that are inconsistent or effect sizes that are small may miss the very truth that genetic epidemiology seeks to discover. At the same time, spurious associations are possible and even likely when hundreds of genotypes can be determined for a subject with only a single sample of blood or saliva.”

Second, without independent confirmation, the analytic validity of the results may be in question. Current regulations require that results given to patients be performed in a laboratory certified under the Clinical Laboratory Improvement Amendments (CLIA). Because CLIA contains an exemption for “research laboratories that test human specimens but do not report patient specific results for the diagnosis, prevention or treatment of any disease” (42 CFR 493.3(b)(2)), many research laboratories are not so certified. This creates a quandary for investigators if they are expected to offer results. If a researcher discloses an individual’s result specifically in response to a request under the Privacy Act or other applicable law (see below), that is generally not considered a “report for diagnosis” because it is being disclosed to comply with the law and not for the medical purposes set out under the exemption. However, the quandary remains with regard to any routine expectation that research results be offered when the research lab is not CLIA certified.

Third, when research involves existing biological materials and consent has been waived, offering results is especially problematic. Genetic test results should never be given to a participant who does not want them [44]. Results should not be returned unless a process of informed consent is in place that includes the opportunity for an informed decision not to receive results.

Finally, creating an ethical or legal obligation to provide research results to participants could confuse the role of the researcher, especially if the researcher is not a physician [31]. Physicians are obligated to act in the best interests of their patients. To the extent that generalizable knowledge is generated and consequently the standard of care is or may be altered, the researcher’s “obligation” to participants--to conduct good science and to disseminate his or her findings widely--is satisfied [44].

According to NBAC, disclosure to individuals the results of research on their biological materials should be an exceptional circumstance and should occur only when all of the following apply (NBAC Rec. 14):

• the findings are scientifically valid and confirmed,
• the findings have significant implications for the participant’s health concerns, and
• a course of action to ameliorate or treat these concerns is readily available.

NBAC further recommends that the investigator in his or her research protocol should describe anticipated research findings and situations that might lead to a decision to disclose the findings to participants, as well as plans for how to manage such disclosure (NBAC Rec. 15). They also recommend that when research results are disclosed to a participant, appropriate medical advice or referral should be provided (NBAC Rec. 16).

Attempting to define the exceptional circumstances under which an after-the-fact determination to offer results might be made could prove problematic. Informing participants about these exceptions would be difficult, and researchers, IRBs, and participants are apt to disagree about what constitutes a finding sufficiently certain or significant to merit disclosure [44]. One alternative is for investigators to communicate at the outset their plans for informing or not informing participants of their results, and then not deviate from that position [53]. These plans can be developed by assessing at the beginning of the research project the likelihood that clinically relevant information will result, based on existing evidence and the aims of the study [44]. Merz et al. [30] state that “when information of potential clinical relevance is a possible result of research activities, then a decision must be made prior to performing the research about whether to de-link the tissues for study or to secure detailed informed consent from subjects, specifically addressing the disclosure and use of such information.” Similarly, Fuller et al. [50] recommend:

If a policy mandating return of clinically meaningless data were implemented, associated costs and personnel might provide an obstacle to doing the study. Thus, in the absence of clinical validity there should not be an absolute requirement for data to be returned to subjects. For research in which data are not provided to subjects, the researcher should demonstrate absence of clinical validity, an IRB should be required to review and approve the exception, and the informed consent document should state explicitly that the data will not be returned to the research subject.

When individual research results are not disclosed, participants could be given the option to receive an aggregate report of overall study results [14]. In the rare event that results unexpectedly have clinical significance, participants could still receive through the aggregate report any recommendation to consider testing for a particular trait in a clinical laboratory, without revealing individual results. In addition, participants who consent to storage and use of their biological sample for future research could be given the option to receive aggregate reports about studies conducted using tissue from the “bank” where their sample is stored [44]. The challenge will be to find ways of presenting research findings in lay language and to be clear about any clinical implications and their meanings in different populations.

This approach may need to be modified in certain instances to meet applicable laws. For example, when a study is subject to the Privacy Act, the Act provides individuals the right to review and get copies of their information (5 USC 552(a)(d)(1)). The Privacy Act applies only when records are maintained by a federal agency in a system of records. The term “system of records” refers to a group of records under the control of a federal agency from which information is retrieved by the name of the individual, identifying number, or some other identifying particular. When research is subject to the Privacy Act, it would be appropriate to inform individuals in advance of their right to see their information. However, although the Privacy Act permits an individual access to his or her records upon request, it does not command affirmative steps to disclose results absent such a request.

Commercialization
Some genetic studies may have the potential to result in commercial gain, for example from gene patents and genetic tests, and thus may create situations in which researchers have interests in conflict with those of research participants. In Moore v. Regents of the University of California [54], a patient suffering from hairy cell leukemia consented to have his spleen removed to slow progress of the disease. The patient’s physician was later awarded a patent that included claims on a cell line derived from the patient’s spleen. The patient filed suit, alleging, among other things, conversion (deprivation of another’s property without his authorization) and lack of informed consent. The California Supreme Court held that the use of excised human cells in medical research does not amount to conversion and that the plaintiff had no rights in the patent. However, the Court noted that when soliciting consent a physician has a fiduciary duty to disclose all information material to the patient’s decision and concluded that this disclosure must include personal interests unrelated to the patient’s health, whether research or economic, that may affect the physician’s medical judgment.

Although the potential for commercial gain may be remote in many population-based studies involving genetics, some participants may object to supporting any such gain or may be opposed in principle to the use of patents to secure intellectual property rights related to the human genome [45]. Thus, any possibility of commercial gain should be disclosed in consent documents, along with a statement about whether participants would share in any profits [13].

Conclusions
The ability to use genetic information to benefit human health requires population-based data concerning the prevalence of gene variants, their associations with disease, and their interactions with modifiable risk factors [55]. As gene discovery continues apace, effective interventions based on knowledge of gene-environment interaction must follow close behind. Without interventions, the risks of genetic information could outweigh the potential benefits for individuals and society, undermining the ultimate value of the Human Genome Project and public trust in the research enterprise.

Recommendations for protecting participants in genetic research have evolved in the context of studies on highly penetrant mutations in families with a heavy burden of disease. These recommendations are based on justifiable concerns, which all epidemiologists would share, about the misuse of highly predictive genotypes [13]. However, the risks associated with population-based research on lower-penetrance gene variants and their interactions with environmental exposures may oftentimes be lower in probability and magnitude than those associated with family studies.

This chapter highlighted some of the ethical, legal, and social issues relevant to the design and conduct of human genome epidemiologic research. It explored these issues using ethical principles, federal policy for the protection of human research participants, and guidelines for research involving human tissue, particularly NBAC’s recommendations. In its report on human biological materials, NBAC confined its recommendations to interpretations of existing federal regulations, but subsequently undertook a comprehensive assessment of the basic purpose, structure, and implementation of research oversight and recommended broad, strategic changes to the oversight system [56]. Regardless of any regulatory or procedural changes that may occur, however, the same ethical principles will apply. Epidemiologists, genetics professionals, policy makers, and the general public can work together to develop recommendations for the protection of participants in population-based research that involves genetics. By protecting participants from the risks specific to such research, we promote our ability to fulfill the promise of genetic information to improve health and prevent disease.

Acknowledgments
Thanks to Dr. Muin Khoury and Dr. Marta Gwinn for their contributions to earlier versions of this paper. This project was supported under a cooperative agreement from the Centers for Disease Control and Prevention through the Association of Teachers of Preventive Medicine.

1. Beauchamp TL. Moral foundations. In: Ethics and Epidemiology. Coughlin SS, Beauchamp TL, eds. New York: Oxford University Press, 1996; 24-52.
2. Gostin LO, Hodge JG. Genetic privacy and the law: an end to genetics exceptionalism. Jurimetrics Journal 1999; 40: 21-58.
3. Rothenberg KH. Breast cancer, the genetic “quick fix,” and the Jewish community. Health Matrix Journal of Law-Medicine 1997; 7: 97-124.
4. Andrews LB, Fullarton JE, Holtzman NA, Motulsky AG (eds). Assessing Genetic Risks: Implications for Health and Social Policy. Washington DC: National Academy Press, 1994.
5. Holtzman NA, Watson MS (eds). Promoting Safe and Effective Genetic Testing in the United States: Final Report of the Task Force on Genetic Testing. Baltimore MD: Johns Hopkins University Press, 1998.
6. Fayerweather WE, Higginson J, Beauchamp TL (eds). Ethics in epidemiology. J Clin Epidemiol 1991; 44: Suppl I.
7. Council for International Organizations of Medical Sciences. International guidelines for ethical review of epidemiological studies. Law Med Health Care 1991; 19: 247-258.
8. World Health Organization. Ethical issues in genetic research, screening and testing with particular reference to developing countries. In Genomics and World Health, Geneva, WHO, 2002; pp. 147-173.
9. National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research. Washington DC: US Government Printing Office, 1979.
10. American Society of Human Genetics. Statement on informed consent for genetic research. Am J Hum Genet 1996; 59: 471-474.
11. Coughlin SS, Khoury MJ, Steinberg KK. BRCA1 and BRCA2 gene mutations and risk of breast cancer. Public health perspectives. Am J Prev Med 1999;16: 91-98.
12. Collins FS, McKusick VA. Implications of the Human Genome Project for medical science. JAMA 2001; 285: 540-544.
13. Hunter D, Caporaso N. Informed consent in epidemiologic studies involving genetic markers. Epidemiology 1997; 8: 596-599.
14. Clayton EW, Steinberg KK, Khoury MJ, Thomson E, Andrews L, Kahn MJ, Kopelman LM, Weiss JO. Informed consent for genetic research on stored tissue samples. JAMA 1995; 274: 1786-1792.
15. McEwen JE, Reilly PR. Stored Guthrie cards as DNA “banks.” Am J Hum Genet 1994; 55: 196-200.
16. National Center for Health Statistics. NHANES III DNA Specimens Guidelines for Proposals. Jun 26, 2001. Available at http://www.cdc.gov/nchs/about/major/nhanes/dnafnlgm2.htm. Accessed Aug 24, 2002.
17. Voelker R. Two generations of data aid Framingham's focus on genes. JAMA 1998; 279: 1245-1246.
18. Mitka M. New heart study a legacy for the future. JAMA 2000; 283: 38, 41, 44.
19. McNicholl JM, Cuenco KT. Host genes and infectious diseases: HIV, other pathogens, and a public health perspective. Am J Prev Med 1999; 16: 141-154.
20. American College of Medical Genetics Storage of Genetics Materials Committee. Statement on storage and use of genetic materials. Am J Hum Genet 1995; 57: 1499-1500.
21. Office for Protection from Research Risks. Issues to Consider in the Research Use of Stored Data or Tissues. Nov 1997. Available at http://ohrp.osophs.dhhs.gov/humansubjects/guidance/reposit.htm Accessed Aug 24, 2002.
22. Grizzle W, Grody WW, Noll WW, Sobel ME, Stass SA, Trainer T, Travers H, Weedn V, Woodruff K. Recommended policies for uses of human tissue in research, education, and quality control. Ad Hoc Committee on Stored Tissue, College of American Pathologists. Arch Pathol Lab Med 1999; 123: 296-300.
23. NBAC. Research Involving Human Biological Materials: Ethical Issues and Policy Guidance--Vol. I. Rockville, MD: National Bioethics Advisory Commission, 1999.
24. de Andrade M, Thandi I, Brown S, Gotto AJr, Patsch W, Boerwinkle E. Relationship of the apolipoprotein E polymorphism with carotid artery atherosclerosis. Am J Hum Genet 1995; 56: 1379-1390.
25. Caggana M, Walker K, Reilly AA, Conroy JM, Duva S, Walsh AC. Population-based studies reveal differences in the allelic frequencies of two functionally significant human interleukin-4 receptor polymorphisms in several ethnic groups. Genet Med 1999; 1: 267-271.
26. Merz JF. IRB review: necessary, nice, or needless? Ann Epidemiol 1998; 8: 479-481.
27. Buchanan A: An ethical framework for biological samples policy. In: Research Involving Human Biological Materials: Ethical Issues and Policy Guidance--Vol. II. Rockville, MD: National Bioethics Advisory Commission, 2000.
28. Human Genome Organisation Ethics Committee. Statement on DNA sampling: control and access. 1998. Available at http://www.hugo-international.org/hugo/sampling.html. Accessed Aug 24, 2002.
29. Outbreak of acute febrile illness among athletes participating in triathlons--Wisconsin and Illinois, 1998. MMWR 1998; 47: 585-588.
30. Update: leptospirosis and unexplained acute febrile illness among athletes participating in triathlons--Illinois and Wisconsin, 1998. MMWR 1998; 47: 673-676.
31. Merz JF, Sankar P, Taube SE, Livolsi V. Use of human tissues in research: clarifying clinician and researcher roles and information flows. J Investigative Med 1997; 45: 252-257.
32. Office for Human Research Protections. Protecting Human Research Subjects: Institutional Review Board Guidebook. 1993. Available at http://ohrp.osophs.dhhs.gov/irb/irb_guidebook.htm. Accessed Aug 24, 2002.
33. Hennekens CH, Buring JE. Epidemiology in Medicine. Mayrent SL, ed. Boston MA: Little, Brown & Co. 1987.
34. Steinberg KK, Sanderlin KC, Ou CY, Hannon WH, McQuillan GM, Sampson EJ. DNA banking in epidemiologic studies. Epidemiol Rev 1997; 19: 156-162.
35. Bell J. The new genetics in clinical practice. BMJ 1998; 316: 618-620.
36. Capron AM. Protection of research subjects: do special rules apply in epidemiology? J Clin Epidemiol 1991; 44: 81S-89S.
37. NCHS. Staff Manual on Confidentiality. Hyattsville, MD: National Center for Health Statistics; DHHS Publ No. (PHS) 84-1244, 1984.
38. Hanson RL, Ehm MG, Pettitt DJ, Prochazka M, Thompson DB, Timberlake D, Foroud T, Kobes S, Baier L, Burns DK, Almasy L, Blangero J, Garvey WT, Bennett PH, Knowler WC. An autosomal genomic scan for loci linked to type II diabetes mellitus and body-mass index in Pima Indians. Am J Hum Genet 1998; 63: 1130-1138.
39. Horgan J. The End of Science: Facing the Limits of Knowledge in the Twilight of the Scientific Age. New York: Broadway Books, 1997.
40. Foster MW, Eisenbraun AJ, Carter TH. Communal discourse as a supplement to informed consent for genetic research. Nat Genet 1997; 17: 277-279.
41. Foster MW, Bernsten D, Carter TH. A model agreement for genetic research in socially identifiable populations. Am J Hum Genet 1998; 63: 696-702.
42. Juengst ET. Groups as gatekeepers to genomic research: conceptually confusing, morally hazardous, and practically useless. Kennedy Inst Ethics J 1998; 8: 183.
43. Reilly PR. Rethinking risks to human subjects in genetic research. Am J Hum Genet. 1998; 63: 682-685.
44. Beskow LM , Burke W, Merz JF, Barr PA, Terry, S, Penchaszadeh VB, Gostin LO, Gwinn M, Khoury MJ. Informed consent for population-based research involving genetics. JAMA. 2001; 286: 2315-2321.
45. Reilly PR, Boshar MF, Holtzman SH. Ethical issues in genetic research: disclosure and informed consent. Nat Genet 1997; 15: 16-20.
46. Woodward B. Challenges to human subject protections in US medical research. JAMA 1999; 282: 1947-1952.
47. Yang Q, Khoury MJ: Evolving methods in genetic epidemiology. III. Gene-environment interaction in epidemiologic research. Epidemiol Rev 1997; 19: 33-43.
48. Centers for Disease Control and Prevention. Consent for CDC research: a reference for developing consent forms and oral scripts. Nov 1998. Available at http://www.cdc.gov/od/ads/hsrconsent.pdf. Accessed Aug 24, 2002.
49. Wertz DC: Archived specimens: a platform for discussion. Community Genet 1999; 2: 51-60.
50. Feinleib M. The epidemiologist's responsibilities to study participants. J Clin Epidemiol 1991; 44: 73S-79S.
51. Fuller BP, Ellis Kahn MJ, Barr PA, Biesecker L, Crowley E, Garber J, Mansoura MK, Murphy P, Murray J, Phillips J, Rothenberg K, Rothstein M, Stopfer J, Swergold G, Weber B, Collins FS, Hudson KL. Privacy in genetics research. Science 1999; 285: 1360-1361.
52. Armstrong K. Genetic susceptibility to breast cancer: from the roll of the dice to the hand women were dealt. JAMA 2001; 285; 2907-2909.
53. Holtzman NA, Andrews LB. Ethical and legal issues in genetic epidemiology. Epidemiol Rev 1997; 19: 163-174.
54. Moore v. Regents of the University of California. 1990; 793 P.2d 479.
55. Khoury MJ, Dorman JS. The Human Genome Epidemiology Network (HuGE Net). Am J Epidemiol 1998; 148: 1-3.
56. NBAC. Ethical and Policy Issues in Research Involving Human Participants--Vol. I. Bethesda, MD: National Bioethics Advisory Commission; 2001.