U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Academies of Sciences, Engineering, and Medicine; National Academy of Medicine; National Academy of Sciences; Committee on Human Gene Editing: Scientific, Medical, and Ethical Considerations. Human Genome Editing: Science, Ethics, and Governance. Washington (DC): National Academies Press (US); 2017 Feb 14.

Cover of Human Genome Editing

Human Genome Editing: Science, Ethics, and Governance.

Show details


Somatic gene and cell therapies are widely seen as morally acceptable. Indeed, bone marrow transplantation, in which cells of differing genetic composition are introduced into patients, has been used for decades, and the use of gene therapy to treat children with severe combined immune deficiency, so-called bubble boy disease, is now a part of medical treatment (De Ravin et al., 2016). Beyond issues of safety, efficacy, and informed consent, there has been no concern about the legitimacy of somatic cell and gene therapies among those who generally endorse modern medicine. Genome editing is playing an increasing part in somatic gene therapy to treat and prevent disease (see Chapter 4). Recent advances, however, have increased the possibility that genome editing could also be used for purposes that go beyond the kinds of gene therapy and other medical interventions discussed above. Thus, the question has been raised anew as to whether enhancement should be regulated or prohibited, and whether there are important differences depending on whether the enhancements are somatic or heritable.

This chapter explores the possible applications of genome editing to achieve what is commonly described as “enhancement,” a term that itself is problematic; it implies a change from—indeed an improvement upon—an existing condition. Enhancements may range from the mundane, such as cosmetic changes in hair color; to the more physically interventionist, such as elective cosmetic surgery; to the more dangerous and problematic, such as the use of some steroids and other drugs among athletes in competitive settings.

Enhancement is commonly understood to refer to changes that alter what is “normal,” whether for humans as a whole or for a particular individual prior to enhancement. The question of what is meant by normal then arises. Is it average? Is it whatever nature has prescribed? Is it whatever luck has wrought? Given the wide range of capabilities exhibited by humans for any particular trait, there is little basis for deeming any one condition normal or any meaningful value in determining an average. Nonetheless, there have been some attempts to describe the range of conditions that, given the right environment, are consistent with an ability to appreciate life and participate in the world.

This chapter begins by reviewing several key issues in genome editing for enhancement. It then addresses in turn somatic (nonheritable) and germline (heritable) genome editing for enhancement purposes. The chapter ends with conclusions and recommendations.


Before beginning to discuss so-called enhancement, it is important to explore how terminology related to human gene therapy and genome editing has the potential to bias judgments unconsciously. Many diseases are associated with DNA variants, and the common term used to describe a variant—“mutation”—has therefore taken on a negative connotation in common parlance. A distinction often is made between “normal” and “mutant” or “disease (causing)” genes, with the latter being viewed negatively. The term “normal” also is applied to phenotype, or the individual traits that result from interaction between the genotype and the environment. Here it is important to note that the “normal” distribution of any given trait (e.g., height, weight, strength, aural or visual acuity) covers a wide spectrum and can be affected by many factors, including but by no means limited to gene variants, which often interact both with each other and with environmental factors. Hence, the term “normal” denotes a range or spectrum, not some ideal state.

The word “natural” has similarly taken on a positive connotation reflecting a common view that nature produces things that are healthier and generally better than anything artificial—this despite evidence demonstrating that “natural” things can be either safe or intrinsically dangerous. In the present context, genetic variants that exist in nature may either support health or cause disease, and the human population contains multiple variants of most genes (see Chapter 4). Thus, there is no single “normal” human genome sequence; rather, there are multiple variant human genomic sequences (IGSR, 2016), all of which occur in the worldwide human gene pool and, in that sense, are “natural,” and all of which can be either advantageous or disadvantageous.

At any given position in the genome, some variants are more common than others. Some are beneficial and some detrimental, their effects at times depending on such factors as whether a person has one copy (heterozygosity) or two copies (homozygosity) of the variant, or whether the gene is sex chromosome–linked (hemizygous for a gene on the Y chromosome and either heterozygous or homozygous on the X chromosome). Another factor may be the particular environment in which a person lives. A well-known example is the case of sickle-cell anemia. Hemoglobin is the protein that carries oxygen in red blood cells. The most widespread variant encodes a fully functional protein, whereas the sickle-cell variant can cause the protein to aggregate and distort the red cells into a sickle shape if both copies of the gene are this variant (the homozygous state), which in turn causes the symptoms of sickle-cell disease. As noted in Chapter 5, however, being heterozygous for this variant confers some resistance to malaria, and for this reason, the sickle-cell variant has been maintained by natural selection in populations from malaria-prone areas (e.g., Africa, India, and the Mediterranean), the disadvantages of sickle-cell disease at the population level being balanced against the population-level advantages of resistance to malaria. So in this case, which natural variant is advantageous depends on the environment.

In this report, the committee uses the term “variant” and eschews to the extent possible the use of “mutant” or “normal” in referring to gene variants. There is, however, a distinction worth keeping in mind. Many variants are “natural,” and changing a gene variant that is associated with disease, such as the sickle-cell variant of hemoglobin, to a variant that is prevalent in the population (i.e., “natural”) but not disease-causing can be viewed as a therapeutic or preventive change. It is also possible, however, to envision the possibility of changing a gene to a variant form that does not exist (or is rare) in the human gene pool but has some property that could be viewed as an “enhancement” since it is predicted to have a beneficial effect. Such a change is a more radical step than that of replacing a disease-causing variant with a common human variant known not to cause disease.


Personal improvements take many forms. They can require significant personal effort, as in taking piano lessons, or they can be largely independent of personal effort, as in wearing teeth-straightening braces. They can be temporary, as in benefiting from the caffeine in morning coffee, or long-lasting, as in immunization against disease. They can be easily reversible, as in hair coloring, or reversible only with difficulty, as in cosmetic surgery. And they can be provided in connection with a corrective intervention, as in removing cataracts and inserting lenses that provide greater acuity than the person ever had naturally. All of these factors influence how improvements are evaluated in terms of fairness and public acceptability.

Although surveys indicate significant support for gene therapy and genetic engineering to improve the health of both existing individuals and unborn children (see Table 6-1), the possibility of “enhancement” in new and potentially more wide-ranging ways can engender anxiety as well as enthusiasm. In 2016, a Pew study of surveys of more than 4,000 individuals revealed that anxiety outpaced enthusiasm not only for enhancement through somatic genome editing, but also for mechanical and transplant-related enhancement (Pew Research Center, 2016). A single study is not definitive, and public opinion on novel interventions in some other controversial areas (such as in vitro fertilization) has become more favorable over time and with evidence of successes. But the Pew study and many others suggest that policy in this area needs to be developed with full attention to public attitudes and understandings.

TABLE 6-1. Summary of Public Attitudes Toward Aspects of Gene Therapy or Genome Editing as Revealed by a Selection of Surveys.


Summary of Public Attitudes Toward Aspects of Gene Therapy or Genome Editing as Revealed by a Selection of Surveys.

Sometimes, the lines between therapy, prevention, and enhancement are blurred, and even the definition of a “disease” that is to be cured or prevented can be open to debate. For this reason, the distinctions between preventing or treating disease and disability (i.e., “therapy”) and the notion of “enhancement” may not fully capture either public attitudes or public policy options. To the extent that there is any public disquiet about the use of gene therapy for disease prevention as opposed to treatment, it appears to be linked to more generalized concerns about “meddling with nature” or “crossing a line we should not cross” (Macer et al., 1995) (see Chapter 5). This is even more true for interventions that appear unrelated to either disease treatment or prevention. As noted above, Americans appear to be largely unenthusiastic about the idea of “enhancement” (Blendon et al., 2016; Pew Research Center, 2016; see Table 6-1).

It is possible that this lack of enthusiasm is due in part to hesitation concerning innovation, which has been shown to be a common phenomenon throughout history, ranging from things now considered quite ordinary, such as coffee and refrigeration, to things still hotly debated, such as transgenic crops (Juma, 2016). Resistance or skepticism may be an outgrowth of concerns about the degree to which an innovation affects cultural identity or may distort socioeconomic patterns in a fashion that is harmful to at least some part of the population. If and when these concerns are either addressed through remedial measures or shown to be unwarranted, innovations that are needed or perceived as desirable become widely accepted.

What is unclear is whether genome editing for enhancement would follow such a pattern or would be such a disruptive application of a new technology that the resistance would persist over time, or whether new concerns will arise as the technology progressed and new applications emerges. “Status quo bias” is a phenomenon in which the preference for what is familiar can affect the way people form judgments about the merits of an innovation. The predisposition toward the status quo may arise from concerns about transition costs (i.e., how people adapt to the circumstances arising from an innovation), about risk (with innovations assumed to have risks that are less amenable to measurement relative to the status quo), about deviation from what is natural (people holding the unwarranted belief that the past processes of natural evolution have optimized humans for the current environment), and about effects on individuals (concern that technology will diminish the quality of relationships between people).

A means of testing for whether status quo bias is affecting the evaluation of new technology has been suggested. In this “reversal test,” one asks, for example, whether those who think people should not have more influence over their traits would also think it would be good if people had less influence (Bostrom and Ord, 2006). The test is intended to distinguish concerns about an innovation itself from concerns about any move away from the status quo. It can be useful when juxtaposed with arguments about a “slippery slope” (see Chapter 5) because it helps distinguish concerns about the technology as it is used today from concerns about future unwanted extensions of the technology.


Given the wide range of other interventions people undergo to alter their bodies and their personal circumstances, any discussion of so-called enhancement must begin with a working definition. Enhancement has been variously defined as “boosting our capabilities beyond the species-typical level or statistically normal range of functioning,” (Daniels, 2000 in NSF, 2010, p. 3) “a nontherapeutic intervention intended to improve or extend a human trait,” (NSF, 2010) or “improvements in the capacities of existing individuals or future generations” (NSF, 2010; President's Council on Bioethics, 2002, p. 1). One definition focuses on interventions that improve bodily condition or function beyond what is needed to restore or sustain health (Parens, 1998). This is a definition that addresses intent as much as the technical intervention, as most interventions can be used either to “enhance” or to restore. For example, under this definition, improving musculature for patients with muscular dystrophy would be restorative, whereas doing so for individuals with no known pathology and average capability would be considered enhancement. Recognizing the importance of intent as an aspect of “enhancement” is helpful, as most biomedical interventions will be subject (in the United States) to regulation by the U.S. Food and Drug Administration (FDA), whose statutory authority explicitly links the initial risk-benefit balance needed for approval to the “intended” use of the product, even though postapproval uses can range beyond that original intended purpose.

Another definitional matter concerns the meaning of “therapy.” It is understood to encompass treatment of a disease (the definition of which is itself subject to debate, as discussed in the section on fairness and enhancement below). But prevention of disease also often is viewed as being encompassed by therapy. To reduce the risk of breast cancer in a person of average health and with nondeleterious variants, for example, is not to cure a disease or even to prevent one that is likely to occur, but this boosted resistance is often viewed as therapeutic prevention, akin to immunization against infectious diseases. A similar point pertains with respect to reducing cholesterol in persons at no more than average risk for heart disease; this is already a widespread practice in American society, and statins and aspirin are widely used to enhance resistance to heart disease pharmacologically even for those not at high risk. Genome edits for such purposes could be similarly health-promoting.

Box 6-1 summarizes key efforts to delineate the distinctions between therapy and enhancement.

Box Icon

BOX 6-1

Making Distinctions.

Somatic (Nonheritable) Genome Editing, Fairness, and Enhancement

With these distinctions in mind, there appears to be broad international consensus, derived from decades of research and clinical trials for gene therapy, that a somatic intervention undertaken to modify a person's genetic makeup for purposes of treating disease is not only permissible but encouraged, provided it proves to be safe and effective.

Before the modern tools needed to modify DNA were developed, government-supported research was focused on developing solid-organ transplantation to replace damaged or diseased organs and on bone marrow transplantation and reconstitution to cure leukemia and other life-threatening disorders, even though these treatments required substituting donor DNA for the patient's DNA in the solid-organ or blood-forming cells. Those cases fell clearly under what is typically considered medical care. Government-supported research also has been conducted in many countries to advance the fields of gene therapy and regenerative medicine and, more recently, human genome editing to modify the DNA in the blood-forming cells of patients with sickle-cell disease and other blood disorders and some forms of cancer. These precedents, and many others like them, have built robust scientific, regulatory, and ethical oversight structures (see Chapter 4).

As noted above, many discussions of the ethics of enhancement have been based on contrasting the concepts of “therapy” and “enhancement.” However, given the evolution of the role of the physician over the past several decades from a healer of the sick to a promoter of health through preventive measures, the therapy–enhancement duality needs to be modified to accommodate a wide range of preventive interventions, such as vaccines, that are neither therapy nor enhancement but blend into each at the edges. For example, while genome editing to lower the cholesterol level of a patient with severe coronary artery disease would likely be viewed as a therapy, and genome editing of a sibling of the patient with high cholesterol who also had other risk factors for coronary artery disease might be viewed as a preventive measure, genome editing to lower the cholesterol of a healthy 21-year-old child of the patient to reduce disease risk below what is average or “normal” in the general population might be viewed as approaching the line between prevention and enhancement. Interventions thus can be viewed as falling on a therapy–prevention–enhancement spectrum, although the boundaries between the three categories are still open to debate and will likely vary with the specifics of the intervention.

With the growth in understanding of the human genome and of which sequence variants are associated with which conditions, the number of traits that could be addressed by genome editing continues to grow. This growing potential again raises the question of what it means to be “normal” and whether deviations from “normality” are really a disease. Everyone would agree that the manifestation of Tay-Sachs disease is not normal and constitutes a disease, but opinions differ as to whether genetically caused deafness should be considered a disease. It is not normal in the sense of being typical or being consistent with the range of capabilities typically associated with the human species, but it can also be associated with membership in a community of persons sharing this characteristic, many of whom reject the notion that deaf people need to be “cured” or otherwise treated to eliminate or circumvent their lack of hearing.

Commentators have noted that the concept of disease is not always objective, but rather can be the result of social agreement influenced by power and prejudice. In the 1950s, for example, homosexuality was considered a disease, and even today it is occasionally the subject of “therapeutic” interventions aimed at “curing” it. In the 1930s, “criminality” was considered a genetic disorder. Some disability activists began to question whether such traits as dwarfism or deafness should be considered diseases instead of variants that enrich human diversity. It is a question that led to discussions about where the line is drawn between normal and pathological, as well as questions about who gets to draw these lines, what authority they have to draw them, which social dimensions are included or excluded, and what provision is made to contest the decisions.

The discovery of variants that simply increase the odds of developing a disease and others that are associated with diseases whose onset is in later life also has blurred the previously bright line demarcating “disease.” Early ethical debate built on language used by the pioneers in human genetics, who referred to “inborn errors,” noting that most traits targeted for modification were called “errors,” as in mistakes from what was supposed to be. The greatest challenge for the normality standard came from some researchers considering what might best be called enhancements for the purpose of relieving disease. This enhancement would not correct errors, but rather instill traits that some lucky minority of humans already have, such as by enhancing immune function or adding cellular receptors to capture cholesterol (Juengst, 1997; Parens, 1998; Walters and Palmer, 1997). Such alterations could be viewed as enhancements or as leveling the playing field for those not fortunate enough to have these traits at birth, and they also complicate the distinction between therapy and enhancement unless one includes prevention as an intermediate concept.

Evolution of the Unfair Social Advantage Demarcation

While both the somatic/germline and disease/enhancement distinctions have been useful, they (like most categories) are imperfect. Some commentators have focused instead on the effect of an intervention and whether that effect is “fair.” Changes not made by personal effort (such as exercise or music practice) but by external forces (such as hair coloring and cosmetic surgery) are understood to have the capacity to generate a social advantage, but it is an advantage within the realm of species-typical attributes. For some, such changes are made purely for pleasure; for others they represent an effort to “normalize” or “even the odds” with those who have the most favored appearances. Externally induced changes that offer more significant or unusual advantages, such as those providing greater muscle mass or more acute vision or obviating the need for sleep, raise questions about the authenticity of the resulting capacity and whether the individual newly endowed with these capabilities is somehow diminished by having failed to earn them. Yet people are born with unearned varying capacities, some markedly superior to the norm, which raises the question of whether and when an advantage becomes “unfair.”

This is a difficult question to answer precisely because of the highly uneven distribution of abilities in the human population. Unless one assigns great importance to fate, it is difficult to tease out enhancements that allow individuals to fairly match the capacities of others from those that are “unnatural,” “abnormal,” or “excessive.” Furthermore, any attempt to relate enhancement to what is “normal” or “average” risks categorizing efforts to combat widespread “normal” but undesirable aspects of life (e.g., age-related declining eyesight, hearing, and mobility) as a form of “enhancement,” with all the pejorative connotations implied by the word.

Society already condones such efforts for many conditions (cataract surgery, hip replacement) using methods other than genome editing. Some respond to inequality in access by favoring interventions that provide more care to more people, and eschew research investment in and insurance coverage for high-cost innovations. This may be a response to economic conditions or a philosophical view of infirmity as a natural part of life, not necessarily in need of every possible measure for treatment and prevention.

Other societies have responded to inequalities that arise from differential access to medical innovations by trying to increase access and insurance coverage, rather than by restricting research or the marketing of new products or technologies. Unpacking such differences requires distinguishing between restrictions on the research itself and decisions about insurance to cover treatments, a topic that in turn requires inquiry into whether insurance is primarily a public good or a privately purchased service.

Even for societies that tend toward expanding access to respond to inequality, a core concern for some is that enhancements are yet another benefit that would accrue primarily to the individual, without benefit to the population as a whole. John Rawls' influential theory of justice emphasizes the idea of equality. He observes that the luck with which someone is born healthy, talented, or in favored social circumstances is neither earned nor deserved. From this he concludes not that all people must be equalized in outcomes but that further distribution of social goods should be designed to account for this initial inequality. This notion leads to so-called equality-based reciprocity, such that inequalities should be tolerated only when they somehow accrue to the population's general advantage, in particular to the advantage of those least well-off (Rawls, 1999).

Some might conclude, therefore, that a problematic enhancement is one that confers a social advantage beyond that which an individual possesses by fate or through personal effort, and that does not benefit the rest of society in any way or undermines the implicit goals of a competition. Using equality of opportunity and societally useful inequality as guides may help distinguish those forms of enhancement that might generally be tolerated (assuming the risks are proportional to the benefits) from those that would be more controversial. Of course, somatic or germline genome editing for enhancement is very unlikely to be the most profound source of inequality in any setting. But those most uncomfortable with using genome editing for enhancement will likely still be concerned regardless of the size of its contribution.

Looking across these themes, one might conclude that enhancement per se is not the focus of concern, but rather the underlying intent and subsequent effect. One response to this concern is to focus on the technologies and applications and to restrict those most likely to be used to unacceptably exacerbate inequalities. A different response is to insist that communities and governments work to make advantageous enhancements available more generally and focus on reducing undesirable inequalities. Within this range of responses lies the choice of governance policy.

Governance of Nonheritable Somatic Editing for Enhancement of the Individual

The governance and ethics of human enhancement have long been the subject of policy reports. Most recently, the U.S. Presidential Commission for the Study of Bioethical Issues focused on human enhancement related to the use of drugs that affect neurological function (Bioethics Commission, 2015), and the European Excellence in Processing Open Cultural Heritage (EPOCH) project summarized the prevailing modes of governance and the roles of academics and bioethicists in these debates, as well as areas of missing evidence needed to identify real possibilities (European Commission, 2012). Earlier efforts include a 2009 report by the U.S. National Science Foundation (NSF) (2010) and a 2003 report by the U.S. President's Council on Bioethics (2003). In the United Kingdom, the Academy of Medical Sciences, British Academy, Royal Academy of Engineering, and The Royal Society came together in 2012 for a policy-focused workshop on emerging technological enhancements that could affect the workplace (AMS et al., 2012). And the French National Advisory Committee on Ethics and the Life Sciences and the Singapore National Bioethics Commission both produced reports in 2013 focused specifically on neuroenhancement (NCECHLS, 2013) and neuroscience research (BAC Singapore, 2013). All of these reports reflect broad input from the medical, bioethical, and academic communities and provide a rich source of information on the concerns that have been raised about enhancements, as well as the profound challenges entailed in clearly delineating the differences among therapy, prevention, and enhancement.

In the United States, governance of enhancement applications of genome editing would fall, as with other gene therapy, to the FDA, the Recombinant DNA Advisory Committee (RAC), institutional biosafety committees (IBCs), and institutional review boards (IRBs), and the legislature (see Chapter 2). The RAC could provide a venue for discussion of somatic enhancement proposals.1 IRBs and the FDA would look at whether the benefits the enhancement might provide to the individual, to science, and to society are reasonable in light of the risks to the individual, to public health, and to environmental safety. But concerns about culture or societal morals, while important, are generally not within an IRB's remit; the regulations state that an IRB “should not consider possible long-range effects of applying knowledge gained in the research (for example, the possible effects of the research on public policy) as among those research risks that fall within the purview of its responsibility” (45 CFR Sec. 46.111(a)(2)).

Thus, if a protocol holds the potential for great benefit to individuals, and those individuals are willing to accept greater risk, the regulator and IRB might agree that the standard of the possible benefits being reasonable in relation to the risks had been met. If the regulator and IRB decide, however, that there are no real benefits of an enhancement—either to the individual or to science—then even a minimal risk is unjustified. As human genome editing improves technologically, there is every reason to believe that the health and safety risks to individuals will diminish. If these risks become de minimis, one might assume that the potential benefits required to justify the risks also will decline. Thus, as the technology improves, its application could extend from serious illnesses, to less serious illnesses, to prevention, and in the long term to enhancement, however defined.

In the United States, it is also important to keep in mind that once a medical product has been approved for a particular purpose and population, the sponsor is limited to marketing it for these “labeled” indications, but individual physicians are free to use their own judgment and prescribe the product for other uses and other populations2 (see Chapter 4). This “off-label” use complicates the question of governance in the United States and in other jurisdictions with similar rules, such as the European Union, because it makes it more difficult to restrict the use of new medical products to those situations that have the best risk/benefit ratios and the general support of the public.

With regard to enhancements, this regulatory scheme has raised concern that some products will be approved for treatment or prevention of disease but then be used off-label for riskier or less well-justified uses. As noted in Chapter 4, however, the specificity of these edited cells will limit the range of off-label uses for unrelated indications far more than is the case with many drugs. While one might imagine a genome-edited cell therapy for muscular dystrophy being of interest to those with healthy muscle tissue who wish to become even stronger, other examples are more difficult to envision. The specificity of edited cells will make such applications less likely for the foreseeable future.

In addition, the FDA has some authority to restrict off-label uses—for example, through requirements for special patient testing or adverse event reporting—and the U.S. Congress can pass legislation to specifically prohibit certain uses, as has been the case for human growth hormone (see Box 6-2). Other jurisdictions have similar powers and choices. Nonetheless, attention to the possible range of off-label uses is necessary, and the need for some control over off-label use can be anticipated.

Box Icon

BOX 6-2

Human Growth Hormone.

In conjunction with formal regulatory processes, a number of other aspects of governance will affect whether and how genome editing is used for enhancement. These include professional guidelines, which influence physician behavior directly and set standards against which that behavior is judged in cases of possible malpractice (Campbell and Glass, 2000; Mello, 2001). Insurers that offer malpractice coverage also can influence the willingness of physicians to offer certain services (Kessler, 2011). In another capacity, insurers play a role by choosing to cover the cost of using approved technologies based in part on the purpose for which they are going to be used and whether the use is necessary or elective.


As noted in Chapter 5, germline genome editing presents the prospect of inducing heritable changes that could affect multiple generations, not just the child who developed from a genome-edited embryo or gametes. In the context of the discussions around enhancement, this prospect may deepen some of the disquiet concerning those applications that are most distant from disease treatment, disease prevention, and correction of significant physical or social disadvantages relative to the norm. This disquiet is influenced not only by the concerns outlined above with respect to somatic genome editing but also by the long and troubling history of eugenics, a history that included coercive measures and even genocide. This history is replete with dogma that creates hierarchies of human quality based on race, religion, national origin, and economic status, and it demonstrates how scientific concepts, such as natural selection, and public welfare measures, such as public hygiene, can be subverted for purposes of cruel and destructive social policies. These considerations lead to the question of whether “enhancement” applications of heritable germline editing should be prohibited entirely or significantly restricted in ways measurably different from those for purely nonheritable somatic editing.


The term “eugenics” was first used in the late 19th century to define the goal of improving the human species by giving “the more suitable races or strains of blood a better chance of prevailing speedily over the less suitable” (Kevles, 1985, p. xiii). The general idea was to create schemes for encouraging people with “good” bloodlines to have more children and those with “bad” bloodlines to have fewer or no children in order to improve the human species. Given their extremely limited understanding of what traits were truly heritable, eugenicists in various societies applied their social biases in ways now deemed unacceptable. In Britain, eugenicists assumed that the “good” traits were those found among the upper classes. They inferred that the fine qualities of the aristocracy were heritable, so the poor should simply produce fewer children. In the United States, the original eugenic impulse involved race or ethnicity. One eugenic goal was to keep races with “bad” traits from immigrating to America. The peak effort in meeting this goal was the 1924 Immigration Control Act, which limited immigrants from Eastern and Southern Europe. It was signed by President Coolidge, who had earlier claimed that “America must be kept American” because “biological laws show that Nordics deteriorate when mixed with other races” (Kevles, 1985, p. 97).

By the 1920s, countries increasingly looked at people's individual qualities independent of their race and class, trying to identify those with supposedly genetic traits such as “feeblemindedness” and “criminality” and discouraging them from reproducing. These eugenics programs were not necessarily voluntary, and many felons and women were forcibly sterilized. Most famously, Justice Oliver Wendell Holmes of the U.S. Supreme Court wrote the opinion that allowed the sterilization of Carrie Buck, a “feeble-minded” woman, concluding that sterilization was justified because “[i]t is better for all the world, if instead of waiting to execute degenerate offspring for crime, or to let them starve for their imbecility, society can prevent those who are manifestly unfit from continuing their kind. The principle that sustains compulsory vaccination is broad enough to cover cutting the Fallopian tubes. Three generations of imbeciles are enough.”3 Eugenics programs were part of progressive social reforms and were thought to uplift the population by improving genetic qualities (Lombardo, 2008).

The logic of eugenics was taken to its extreme conclusion in Nazi Germany, where those perceived to have genetically derived limitations were first sterilized and in later years killed. The logic of eugenic purity was a part of the Holocaust, which resulted in the deaths of millions of people portrayed as genetic inferiors, primarily Jews, Roma, and those with disabilities. According to historian Daniel Kevles, after revelation of the Holocaust, people realized that “a river of blood would eventually run from the German sterilization law of 1933 to Auschwitz and Buchenwald” (Kevles, 1985, p. 118).

Reform Eugenics

Revelation of the Holocaust was not the end of eugenics, only the end of race-based, coercive, state-mandated eugenics. Many scientists had already rejected the mainstream eugenic view, with Hermann Muller writing in 1935 that eugenics had become “hopelessly perverted” into a pseudo-scientific facade for “advocates of race and class prejudice, defenders of vested interests of church and state, Fascists, Hitlerites, and reactionaries generally” (Kevles, 1985, p. 164). Muller and other prominent scientists, such as Julian Huxley, would create a “reform” eugenics that sought to stop the reproduction of people with genetic disease and encourage more reproduction by people with “superior” genes—from whatever race and class. People would be encouraged to change their reproductive practices voluntarily for the good of the species. Most notably for future debates, these thinkers wanted humanity to seize control of its own evolution and improve the species in various ways, such as by making humans more intelligent. The ethical debate of the 1950s through the early 1970s was quite broad, often focused on what the goals for genetic modification as a species should be. In a theme that would recur from this era forward, some critics of reform eugenics averred that humans should be satisfied with the way they are. In general, the ethical debate was about the genetic goals of the species—or whether to have such goals at all (Evans, 2002).

In 1953, Crick and Watson described the structural basis of how DNA duplicates itself (Watson and Crick, 1953), leading to understanding of how the DNA of genes encodes information. This discovery changed the ethical debate concerning eugenics as people realized that if genes were actually chemicals whose structure could be characterized, society no longer would have to rely on “who mates with whom”; rather, people could be chemically modified to have “more” of the “good” genes. Robert Sinsheimer, a prominent scientist of the time, was typical in his response, writing in 1969 that “the old eugenics would have required a continual selection for breeding of the fit, and a culling of the unfit. The new eugenics would permit in principle the conversion of all of the unfit to the highest genetic level . . . for we should have the potential to create new genes and new qualities yet undreamed in the human species” (Sinsheimer, 1969, p. 13). Theologian Paul Ramsey wrote in 1970 that such proposals would make “man” “his own self-creator” and lead to a new theology of science (Ramsey, 1970, p. 144).

Whether to improve the species and, if so, in what way was the core of the ethical debate until the early 1970s. A discussion then—one continuing today among some transhumanists—is whether human evolution should be left to processes of natural selection, which are random and occur very slowly. For example, Corneliu Giurgea, the Romanian chemist who synthesized Piracetam in 1964 and showed that it might act in cognitive enhancement, said, “Man is not going to wait passively for millions of years before evolution offers him a better brain” (Giurgea, 1981). Indeed, with the specter of climate change on earth and the imagined colonization of Mars, some transhumanists today discuss whether humans need to intervene in their own evolution to cope with the future they are creating (Bostrom, 2005; Rosen, 2014). But in the 1970s, the evident complexity of making changes, let alone determining which changes are desirable, led many to rethink what prominent biologist Bernard Davis would dub “Promethean predictions of unlimited control.” He reminded readers of facts now considered obvious, such as that most “enhanced characteristics” are polygenic and thus difficult or impossible to modify (Davis, 1970, p. 1279).

Technological limits of that era also helped shape the debate. It was impossible to imagine somatic enhancements when somatic therapy was not yet successful, so any claim of a somatic enhancement would have been considered too risky, with very little benefit. Similarly, if early attempts at modifying somatic cells through viral vectors were successful in only a small number of the cells, how could sperm, eggs, or zygotes be changed? But where enhancement and heritable change came together, even though not yet technically feasible, public concern was greatest. And the goal of the eugenicists—to make the species better—was placed in this category, thus comingling the rejection of eugenics with the possibilities for germline editing.

Slippery Slope Concerns about Germline Enhancement

Opponents of germline enhancement from the reform eugenics era forward were concerned largely with long-term cultural or social changes that could occur as the result of a sociological slippery slope process (see also Chapter 5). That is, would somatic therapy eventually lead to efforts to enhance the species through germline engineering, a process that might evolve as skill and familiarity with somatic therapy made it easier to imagine other applications as helpful and safe? These opponents were willing to endorse somatic therapy because they thought the somatic/germline distinction was culturally strong, and the public would make a clear distinction between modifying individuals and modifying their offspring (Burgess and Prentice, 2016).

This sort of slippery slope argument emerged in 1981 after a U.S. Supreme Court decision that allowed the patenting of genetically engineered life forms (Evans, 2002). Concerns were raised about “the fundamental nature of human life and the dignity and worth of the individual human being” (President's Commission, 1982, p. 95). The presidential bioethics commission of that era wrote a report entitled Splicing Life, in which it reformulated the ethical debate so that the report would be “meaningful to public policy consideration” (President's Commission, 1982, p. 20). To make ethical claims legally actionable meant moving away from arguments about future cultural harms or claims that it is not the role of humanity to modify itself. Consequences needed to be more concrete and near term, not speculative. In the report, the commission stated that it “could find no ground for concluding that any current or planned forms of genetic engineering, whether using human or nonhuman material, are intrinsically wrong or irreligious per se” (President's Commission, 1982, p. 77). The report established a framework of risks and benefits and the rights of individuals that would serve as a framework for government regulation of this new science, such as through the human gene therapy subcommittee of the RAC. It was an approach not particularly amenable to consideration of broader and longer-term social effects, for instance slippery slopes, because of its focus on more immediate effects on identifiable persons.

The pre-1980s ethics debate returned in 2001 with the appointment of a federal bioethics commission by President George W. Bush. This commission claimed that it “eschewed a thin utilitarian calculus of costs and benefits, or a narrow analysis based only on individual ‘rights.’” Instead, it claimed to ground its reflections “on the broader plane of human procreation and human healing, with their deeper meanings” (President's Council on Bioethics, 2002, p. 10). Most notably, the commission was concerned about the promotion of inequality and about parents having the ultimate power over their children (President's Council on Bioethics, 2003, p. 44). The report did not have a strong impact on policy regarding human genetic modification, but made clear that some in U.S. society viewed germline genetic modification through this particular ethical lens.

The concern of the Bush-era commission that germline enhancement might encourage people to view children as something to be designed and manipulated has long been a concern of some social scientists and humanists. Political theorist Michael Sandel wrote that “to appreciate children as gifts is to accept them as they come, not as objects of our design or products of our will or instruments of our ambition. Parental love is not contingent on the talents and attributes a child happens to have” (Sandel, 2013, p. 349). One implication is that potential parents should refrain from making modifications that would directly benefit their future child, not necessarily because doing so would cause them to see their own child differently, but because it might in a tiny, indirect yet cumulative way promote a culture that would come to see all children differently. Yet critics of this view would argue that the harm is speculative, and that this level of freedom in the relationship of the individual to society is well within the range of what is allowed in liberal democratic societies.

Another concern that has been raised revolves around whether parents might become increasingly viewed as responsible for the qualities of their offspring. According to Sandel, “we attribute less to chance and more to choice. Parents become responsible for choosing, or failing to choose, the right traits for their children” (Sandel, 2004, p. 60).

A similar idea is expressed in the more religious language of making versus begetting. Theologian Gilbert Meilaender describes designing the genetic qualities of one's children as akin to “making” them, whereas nondesign is “begetting” them. More important, and like Sandel, he states that “what we beget is like ourselves. What we make is not; it is the product of our free decision, and its destiny is ours to determine” (Meilaender, 1997, p. 42). By this view, begetting (i.e., nondesign) is critical to human dignity and human rights because “we are equal to each other, whatever our distinctions in excellence of various sorts, precisely because none of us is the ‘maker’ of another one of us” (Meilaender, 2008, p. 264). These concerns about objectification might possibly apply to germline conversion to genes associated with ordinary health, but would more likely be raised by enhancements, health-related or beyond.

There are other views, of course. One might say that making choices about our genetic future—whether or not they increase the perception that humans are more like objects—is precisely human. As Joseph Fletcher, one of the founders of bioethics, wrote in 1971: “Man is a maker and a selecter and a designer, and the more rationally contrived and deliberate anything is, the more human it is. . . . [T]he real difference is between accidental or random reproduction and rationally willed or chosen reproduction” (Fletcher, 1971, pp. 780-781).

Others have argued that parental discretion allows for a wide range of practices, provided they are not significantly harmful to the physical or psychological development of a child (Robertson, 2008). As discussed in Chapter 5, this view requires a pure reproductive rights framework that must be stretched to its limits to include the right to enhance or diminish traits (Robertson, 2004). It is a vision of parental liberty that already encompasses a wide range of enhancements of infants and children, including such biomedical measures as cosmetic surgeries, the use of growth hormone for short stature of unknown cause (see Box 6-2), and the use of some performance-enhancing drugs. By extension, it could be argued that this liberty encompasses germline enhancements with similar risk/benefit balances. Here again, though, there is very little reason to think that in the United States, the constitutional cases on parenting would prevent the government from banning germline genome editing if it had a rational basis for doing so.

Academic transhumanism has emerged as a contributor to these debates. Transhumanists argue not only for the ethical legitimacy of some forms of enhancement-oriented germline editing, but perhaps even for parental responsibility and an ethical obligation to take advantage of such enhancement possibilities for the benefit of one's children (Persson and Savelescu, 2012). One philosopher has argued that humans are obligated to make the best possible decisions for those who cannot decide for themselves, and this would include future children and their descendants (Harris, 2007). These are arguments about moral obligations, however, as nothing in U.S. statutes or judicial decisions (or those of other countries) imposes them as a matter of law.

Overall, two distinct approaches to evaluating the ethics of germline enhancement have emerged over the past half century. In one, the focus is more societal and philosophical. It encompasses not only the concerns raised about germline editing in general, as described in Chapter 5, but also concerns about altering how children are viewed and about creating or increasing social inequities in a multigenerational fashion as a result of the heritability of the enhancement. Even where the benefits of an individual enhancement might be regarded as justification for an individual intervention, these analyses often feature a concern about the slippery slope and an echo of eugenics movements of the past.

In another approach, the disease/enhancement distinction remains largely useful, as it tracks well to the evaluation of individual risks and benefits. This evaluation is the focus of the regulatory bodies, such as the FDA, that review new medical products for approval and the research oversight bodies that oversee protection of clinical trial participants and others who might be put at physical risk by the trials. When diseases are cured or prevented, the benefit of trials is seen as greater relative to when functional traits are improved beyond what is necessary for a typical life. In turn, this gradation of benefits is balanced against health risks for offspring and future generations, including the potential for disease prevention.

Given that human germline genome editing has not yet been tested clinically for therapeutic or preventive purposes, it appears clear that germline genome editing for purposes of enhancement—that is, not clearly intended to cure or combat disease and disability—is very unlikely at this time to meet the standard of possible benefit and tolerable risk as required to initiate clinical trials. Even as risks recede with greater experience and information, truly discretionary and elective germline edits would be unlikely to have benefits outweighing even minor health risks.


Significant scientific progress will be necessary before any genome-editing intervention for indications other than the treatment or prevention of disease or disability can satisfy the risk/benefit standards for initiating a clinical trial. This conclusion holds for both somatic and heritable germline interventions. There is significant public discomfort with the use of genome editing for so-called enhancement of human traits and capacities beyond those typical of adequate health. Therefore, a robust public discussion is needed concerning the values to be placed upon the individual and societal benefits and risks of genome editing for purposes other than treatment or prevention of disease or disability. These discussions would include consideration of the potential for introducing or exacerbating societal inequities, so that these values can be incorporated as appropriate into the risk/benefit assessments that will precede any decision about whether to authorize clinical trials.

RECOMMENDATION 6-1. Regulatory agencies should not at this time authorize clinical trials of somatic or germline genome editing for purposes other than treatment or prevention of disease or disability.

RECOMMENDATION 6-2. Government bodies should encourage public discussion and policy debate regarding governance of somatic human genome editing for purposes other than treatment or prevention of disease or disability.



At the moment, the RAC is not accepting germline editing protocols for review (see Chapters 2 and 5).


59 Federal Register 59, 820, 59, 821 (November 18, 1994). “Once a [drug] product has been approved for marketing, a physician may prescribe it for uses or in treatment regimens of patient populations that are not included in approved labeling.” The notice goes on to state that “unapproved” or, more precisely, “unlabeled” uses may be appropriate and rational in certain circumstances, and may in fact reflect approaches to drug therapy that have been extensively reported in medical literature.


Buck v. Bell, 274 U.S. 200 (1927).

Copyright 2017 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK447264


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (3.9M)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...