Citation:
Date Presented:
04/2009Full Text
The Politics of Genomics Research:
The Implications of DNA for Racial Identity and Race-based Medicine
Jennifer L. Hochschild
and
Maya Sen
Government Department
Harvard University
March 26, 2009
Paper prepared for the annual convention of the Midwest Political Science Association, Chicago IL, April 2-5, 2009.
DRAFT: please do not quote or cite without permission from the authors. We welcome comments and suggestions; please send them to Hochschild@gov.harvard.edu or msen@fas.harvard.edu.
Our deep thanks to Justin Grimmer and Gary King for their invitation to us to be early experimenters with their new methods, and for their patient and helpful advice in conducting the analyses.
ABSTRACT: The extraordinary growth and spread of genomics research and DNA technology will soon, although it does not now, have some sort of partisan association and ideological valence. We plan to research that development, beginning in this paper by examining two issues: the tension between individualized and race-based medical research, diagnosis, and treatment; and the tension between racial boundary blurring and racial boundary strengthening in the social use of DNA for genealogical discovery.
We examine those two issues by using new software to cluster and then to content-analyze a database of all relevant newspapers articles about “DNA and medicine” or “DNA and race” in Lexis-Nexis since 1988, when they began to appear. We find a slight preponderance of articles focusing on individualized rather than race-based medicine. Among articles focusing on non-blacks’ genealogical search, we find a much larger share focusing on racial boundary-blurring rather than reification. But among articles focusing on blacks’ genealogical search, the pattern is somewhat different. We find little change over time, but that may be a methodological artifact rather than a genuine result.
The paper concludes with suggestions about the meaning of these findings, indications of further work to be done, and observations about the importance and uniqueness of the genomics revolution for political science and politics.
The genomics revolution is underway. “There is in biology at the moment a sense of barely contained expectations reminiscent of the physical sciences at the beginning of the 20th century. It is a feeling of advancing into the unknown and [a recognition] that where this advance will lead is both exciting and mysterious…. The analogy between 20th-century physics and 21st-century biology will continue, for both good and ill” (Economist 2007). Even sober and cautious researchers report “a great deal of excitement, because everyone realizes the field is changing so fast” (Science 2007): 821). The effects of this revolution appear widely -- in genetic food modification, breakthroughs in medical research, concerns about insurance companies’ or employers’ use of personal genetic information, courts’ use both to convict and to free those wrongfully convicted, identification of victims of the Srebenica massacre, and “vanity DNA testing” to discover one’s genetic heritage (or that of one’s pet).
Despite the breadth of genomics research[1] and the speed of discovery (or perhaps because of them), the public, policy-makers, and scholars have little information about and even fewer analytic frameworks for understanding genomics. What is the Democratic or Republican, the liberal or conservative, stance on race-based medicine? How will citizens balance exciting genetic discoveries against the risk of claims of finding a “gene for violence” in one race or nationality? How should regulators and pharmacological researchers weigh the need for privacy against the use of new genetic information? What stance, if any, should elected officials and policy-makers take with regard to controlling, supporting, or otherwise interacting with the results of genomics research?
We are beginning a long-term project to study the growth of the ideology and politics of genomics. At present there are few laws, regulations, or court cases; few politicians or policy makers have taken a position on genomics; the public is largely innocent of policy views or ideological associations with the topic. But we predict confidently that a decade from now, those vacuums will be filled. Soon DNA tests will be readily available and cheap; one can imagine that within a generation every newborn will have a genetic profile attached to his or her birth certificate. We want to examine how Americans come to formulate the politics of and attach values to this “advance into the unknown.”
Themes
This paper initiates the larger project by focusing on two themes. The first is the tension between personalized and race-based medicine. On the one hand, doctors will soon be able to identify patients’ genetic information and tailor medical care for some diseases. For example, “cancer doctors at Massachusetts General Hospital plan within a year to read the genetic fingerprints of nearly all new patients’ tumors, a novel strategy designed to customize treatment” (Smith 2009). Doctors will no longer need to remind men perceived to be black to be tested for prostate cancer or women perceived to be Ashkenazi Jews to be tested for the BRCA2 gene. Instead, medical professionals will be able to ignore “race” and focus directly on the patient’s individual characteristics. After all, as geneticists among others have been insisting for a long time, race is (mostly) a social construction; “classification into races has proved to be a futile exercise…. We can identify ‘clusters’ of populations… [but] at no level can clusters be identified with races, since every level of clustering would determine a different partition and there is no biological reason to prefer a particular one” (Cavalli-Sforza et al. 1994): 19).
Individualized medicine carries great promise, a hope that no one denies. Cautions can still be heard, however, especially the concern that doctors focused on a person’s genetic makeup will ignore or undervalue the physiological implications of living in a racially inflected society and polity. Reams of medical literature show the links between disfavored minority status and high blood pressure, persistently high levels of stress, poor environmental conditions, and other social conditions that readily become somatized. If the medical arena becomes too attentive to one’s particular DNA profile, it may miss the social forest for the individual biological tree – or at least that is what many fear.
On the other hand, medicine may become increasingly racialized. In 2005, the FDA approved the drug BiDil for use in treating heart disease for African Americans only. Other race-specific medications are likely to follow as particular diseases become associated with particular groups. For example, “Decode Genetics reported in 2006 that African Americans have twice the incidence of a gene on chromosome 8 that causes susceptibility to prostate cancer, and African American women tend to suffer from very aggressive forms of breast cancer” (Nathan 2007): 12). Or, “reports were published… of a genetic variant associated with heart disease that, in the one-quarter of Caucasians tested who carry two copies, increases the chance of a heart attack by more than 50%” (Science 2007): 821). If particular races are more likely to have particular diseases, then it might make sense to develop race-specific medicines for those diseases, or at least so some argue.
Race-based medicine is likely to prove even more controversial than individualized medicine. Some African American researchers and the Association of Black Cardiologists find this new line of research and medication very promising. Howard University established the National Human Genome Center (NHGC) in 2001 to “explore the science of and teach the knowledge about DNA sequence variation and its interaction with the environment in the causality, prevention, and treatment of diseases common in African American and other African Diaspora populations” (www.genomecenter.howard.edu/intro.htm). Its scientists investigate genes associated with obesity, diabetes, hypertension, and metabolic syndrome, while remaining attentive to interactions between genes and environment. (www.genomecenter.howard.edu/News%20documents/BU_Howard_FINAL_10%5B2%5D.06.pdf). A spokesperson for the Association of Black Cardiologists responded to critics in Scientific American by arguing that “passion directed against this paradigm [“race as a descriptor of drug efficacy”] is misguided, as the consequences of persistent argument and debate unfortunately are persistent disease and premature death” of African American heart patients (Yancy 2007).
Yancy’s description of opponents as “passionate” is accurate. Sociologist Troy Duster warns that “recent research in medicine and genetics makes it… crucial to resist actively the temptation to deploy racial categories as if immutable in nature and society.” Genomics research “holds promise” for finding drugs for “special subpopulations that have some functional genetic markers.” But doctors must not use phenotype or nominal racial categories as a shortcut to identifying those subpopulations; “research [must] be conducted to find the markers that have the actual functional association with drug responsiveness – thus assuring that the drug be approved for everyone with those markers, regardless of their ancestry” (Duster 2005): 1050, 1051). Using race as a shortcut for determining possible susceptibility to a disease, in short, risks both over- and under-inclusion.
Duster’s deepest concern, however, is the possibility of research programs aiming to link “SNP patterns [single nucleotide polymorphisms, the tiny genetic unit underlying the new research breakthroughs] and searches for a biological basis for criminal behavior.” That is, from using race as a shorthand for identifying populations at higher risk of a given disease it is only a small step to using race as a shorthand for identifying populations with propensities to violence or sexual aggressiveness (ibid: 1051). And the race so identified will, he anticipates, be black. Another writer is blunter: “We are ill-prepared to respond to the complex challenges posed by racial arguments bobbing in the unstoppable tide of genetic research” (Kohn 2006). This concern is not mere paranoia. “A team of scientists at the University of Utah has proposed that the unusual pattern of genetic diseases seen among Jews of central or northern European origin… is the result of natural selection for enhanced intellectual ability” (Wade 2005). This is a potentially explosive line of research; as cognitive psychologist Steven Pinker observed in response, “It would be hard to overstate how politically incorrect this paper is” (ibid).
One can see, in short, intense arguments both for and against race-based medicine -- with its hope of new cures and risk of reinventing eugenics[2] – as well as for and against individualized medicine -- with its corresponding hope of new cures and risk of ignoring the medical consequences of living in a racialized society. We plan to analyze the politics, policies, and values of this debate, beginning in this paper.
The second theme is the tension between racial boundary-blurring and racial solidification in society and self-identification. New genetic tests sometimes confound established notions of race and ethnicity. Danny Villarreal, for example, is a Hispanic Texan who believed himself to be of pure Spanish blood. A DNA test showed him to be closely related to Jewish populations in Hungary, Belarus, and Poland. Danny Villarreal is (genetically, at least) an Ashkenazi Jew (Lomax 2005). Such stories of racial boundary blurring have proliferated as more people turn to “vanity” DNA tests to learn about their genetic ancestry; journalists love these stories. Henry Louis Gates discovered that “my mitochondrial DNA, my mother’s mother’s mother’s lineage… [was not] Yoruba, as I fervently hoped…. A number of exact matches turned up, leading straight back to that African Kingdom called Northern Europe, to the genes of (among others) a female Ashkenazi Jew.” (His response: "I have the blues. Can I still have the blues?") (Gates Jr. 2006). Even more dramatically, Wayne Joseph “grew up a black American” with a strong racial identity. He expected his DNA test to “ ‘come back about 70 percent African and 30 percent something else’.” Upon receiving the results, however, “ ‘I was floored’.” He was found to be 57 percent Indo-European, 39 percent Native American, and 4 percent East Asian – no African ancestry at all. “For almost a year, Joseph searched his soul…. Before the test, ‘I was unequivocally black. Now I’m a metaphor for America’ ” (Kalb 2006): 55).
Racial boundary-blurring may enable a person to feel greater empathy or connection to distant others. The red-haired, freckled, apparently Scottish Jack Hitt reports that “I carry the DNA marker found in great abundance among the Fulbe tribe of contemporary Nigeria” and he finds that the link has made him look at news stories about Africa differently (Hitt 2005). But boundary-blurring may instead be deeply disturbing. Consider Lisa Lee, formerly active in the Black Power movement, who discovered that she had no African ancestry on father’s side, “What does this mean; who am I then? For me to have a whole half of my identity to come back and say, ‘Sorry, no African here,’ it doesn’t even matter what the other half says. It just negates it all…. It doesn’t fit, it doesn’t feel right” (Harmon 2006): 18).
Given that many if not most African Americans have some white ancestry, that Hispanics are mestizo more or less by definition, that intermarriage rates among American Indians have been very high for a century, and that Asian Americans increasingly marry whites, one can expect that most Americans typically defined as nonwhite will discover a great deal of mixture in their genetic heritage. So will most whites – certainly if one takes ethnic, nationality, or regional differences seriously and probably even if one focuses only on “race” as conventionally defined. Whether racial boundary-blurring is desirable, on the grounds that it breaks down artificial barriers and connects people across time and place, or undesirable, on the grounds that it is disorienting or encourages members of disfavored minorities to “whiten” themselves, remains a matter of intense debate.
But in this arena as in medicine, the opposite phenomenon is simultaneously occurring. New genetic tests can deepen rather than blur racial or ethnic identity, as people become able to determine their tribal or group ancestry. Consider, for example, Mika Stump. She grew up in foster homes, knowing nothing of her roots except that she is black. “But a DNA test she took recently showed strong similarities between Stump’s genetic code and the Mende and Temne people of Sierrra Leone…. Now, ‘I have a place where I can go back and say, “This is who I am; this is my home.” That’s something I never, ever expected to say’ ” (Willing 2006): xx). This story too is becoming a favorite with journalists, as well as with some scholars and commercial firms. The enterprising director of the Centre for Forensic Investigation at Glasgow Caledonian University states his “intention to have DNA swabbing kits in all the tourist information offices and hotel lobbies across the UK, so people can go and pick up a kit for a few pounds then post it off to us and we will do the DNA tests for them.” For the researchers, that will “contribute to a DNA database of Scottish and Irish clan groups.” For the Scottish tourist industry, “DNA testing will be a draw for ancestral tourists who might want to ‘walk in the footsteps of their ancestors’ ” (Lei 2007). Cruise ships offer luxurious trips to one’s ancestral home, and a few countries offer dual citizenship to people who want a more official connection to their roots.
As in the medical arena, one can see the beginning of an intense ideological and political debate. Should Americans seek to obfuscate artificial racial categories, or should they try to find ancestral groups once thought lost? Once one has the results of a DNA test, how much should and do they matter, and in what ways? Are the testing companies simply reifying artificial categories, or are they providing new evidence for old truths about population variation? Will the alternative trajectories of group boundary-blurring and solidification be politicized, and attached to long-standing racial or ethnic divisions? How will and should the legal system or regulators respond to new forms of biological determinism or racial profiling? We plan to analyze the politics, policies, and values of this debate also, beginning in this paper.
Data
To explore the trajectories of these countervailing movements in both medicine and society, we gathered newspaper articles on the subjects of race, ethnic identity, DNA, and genetics. Unlike that of most media scholars, our purpose was not to explore the influence of the media on the public’s attention, priorities, or attitudes. Instead, we used the media as a lens through which to perceive the initial stages of public discourse on the social and medical relationships between genomics and race. That is, we are using the media as a conduit, albeit an imperfect one, to public opinion.
Using the media in this way is in part a temporary expedient, until these issues are more robustly explored by public opinion surveys and perhaps in elections or referenda.[3] However, we do not mean to be defensive; the media provide as good a window as any into the process by which a new social and political issue is created and spread. Media reports probably represent the main channel of learning for adult American citizens and will continue to do so for a generation, until knowledge about genomics diffuses into K-12 curricula and today’s students become tomorrow’s voters.
In any case, the articles came from Lexis-Nexis’ online database of U.S. newspapers. The papers range from the New York Times and the Washington Post to small regional newspapers such as the Flint Journal or the Dodge County Independent News and a smattering of trade publications, business journals, and law-oriented publications. We used the broadest available database on the grounds that it would best serve as a reflection of how the media are reporting these issues and how people from all walks of life are learning about them.
We conducted the search over this database in July 2008, using three steps in order to capture as many relevant articles as we could. The first search was for any article that mentioned a genetics-related search term (“DNA,” “genetic,” or “genomic”) within ten words of similarly broad medical disease terms (“medicine,” “illness,” or “disease”).[4] This search yielded about 120,000 articles, which ran the gamut from community calendars to TV listings, from high school athletics to scientific reporting on cutting-edge medicine. More recent years had many more articles; 5,600 were published in 1996, compared with 10,343 in 2006.
Two additional searches on the same Lexis-Nexis database were more tailored to the themes of this paper. To identify articles that mentioned the relationship between race or ethnicity and genetics, we used race-specific terms instead of general medicine disease terms.[5] This search resulted in 1,692 articles, from 1972 through 2006. To identify articles that explored the use of genetic information by genealogists and those interested in learning more about their ancestral roots, we used genealogy-type search terms instead of general disease terms.[6] That search yielded 5,972 articles, from 1977 through 2006.[7]
These three databases, cleaned and combined as we explain in the next section, provide the evidence for the analysis below.
Methods and Initial Results
It was not feasible to manually code a database of almost 130,000 articles, even if we were able to eliminate duplicates and what might be the many items irrelevant to our themes. Instead, we first used the “expressed agenda method” being developed and implemented by Justin Grimmer (Grimmer forthcoming 2009) to conduct exploratory content analysis. That helped us to gain traction on the shape and focus of the discourse in the articles and to prune down the database to a more manageable size by eliminating duplications and irrelevancies. We then used a content analysis method being developed and implemented by Dan Hopkins and Gary King (Hopkins and King 2008) to focus attention on the topics of interest -- use of DNA technology in blurring or reifying race and ethnic identity, and use of DNA technology in personalized or race-based medicine.
Expressed Agenda Method: Our search terms yielded many articles that were irrelevant or only tangentially related to the themes of this paper. Examples include several hundred articles on sporting events (a baseball player has pitching “in his genes”), animal or plant genetics (the zebra mussel is a freshwater bivalve that rapidly adapts to new environs), and calendar or community events (genealogy discussions and public library talks). The expressed agenda method helped to eliminate them.[8]
We first classified all of the articles in the combined database by the year of publication according to Lexis-Nexis. Punctuation marks and capitalizations were removed and a Porter stemming algorithm transformed words into stems. Next, a Bayesian hierarchical model was used to “probabilistically determine each of the… topics” that appear in each article (Grimmer 2009): XXX). The result is a kind of cluster analysis that systematically groups articles according to key words and concepts identified by the program, after a person has determined how many clusters the program should create. Appendix A1 shows the 45 clusters created by the expressed agenda method in one iteration (after some irrelevant clusters had already been eliminated). We deleted the eleven shaded clusters, on the grounds that they were demonstrably irrelevant to our two themes, and reran the program to create the final database of 34 clusters, shown in Appendix A2. The final database includes 48,092 articles.
The clusters nicely reveal the content and relative proportions of the articles in the database. Discussion of heart disease is the most common topic, bringing together about six percent of the items. Other topics include Alzheimer’s disease (5.4 percent in one cluster, 2 percent in another), breast cancer (4.6 percent), and Thomas Jefferson’s progeny (0.7 percent). Articles connecting DNA research to race or ethnicity are rare. The cluster of “human, dna, year, ancestor, modern, Africa” includes 2.6 percent of the articles, and the “dna, africa, famil, ancestri,… black” category includes another 1.3 percent. Other clusters are at least suggestive of our themes – one with sickle-cell anemia includes 2.9 percent of the items, and another with “heart, diabetes, [blood] pressure, and black” covers 1.5 percent.
The expressed agenda method shows the range of topics, but unless one specifies many more clusters, it will not pick out a brief mention of the racial or ethnic association of a disease or ailment. (For example, many articles discussing breast cancer pointed out that a woman’s race or ethnic background could be correlated with her risk of developing the disease.) We therefore turned to another type of content analyses for a more focused exploration of how the media link DNA research and race.
Content Analysis: Hopkins and King’s software is intended to expedite the tedious process of content analysis, at least for fairly simple categorizations. A researcher first sorts a subset of the articles into several mutually exclusive categories. At least 100 articles should be coded, and there is little improvement in accuracy after 500 are coded. Content-analysis software[9] then uses the coded subset of articles to sort the remaining articles; it predicts with considerable accuracy the proportion of articles that fall into each category. We hand-coded two subsets of the final database in order to determine what the media report about how and why individuals of different ethnic backgrounds use DNA technology, and whether medical research is moving toward individualized or race-based treatments, or both.
For the social DNA testing, we hand-coded 296 articles, chosen non-randomly to reflect a cross section of possible types of articles. In addition to exploring the basic division between boundary-blurring and boundary strengthening, we thought it probable that African Americans would be mainly focused on the latter and other Americans relatively more focused on the former. Non-black Americans have what might be termed the luxury of curiosity about their heritage and openness to surprising results, since their status as white or at least as non-black will not be called into question no matter the results. African Americans, however, might find non-black ancestry to be psychologically fraught given our nation’s history of white male domination over black women. And blacks may have a much more urgent desire to trace their ancestry in response to the destruction of family, community, and place created by slavery and the Middle Passage. In addition, since we are analyzing media reports of, not direct experience with, genealogical DNA testing, it is plausible that reporters find stories about black ancestry more dramatic and newsworthy than stories about non-black ancestry.
These probabilities do not quite rise to the level of a formal hypothesis. Nevertheless, we explored them by placing each article in one of the following categories:
1. Non-black blend: The article implied or stated that recent advances in genetics or DNA technology blend racial identity, and their subject is implicitly or explicitly non-black. For example, (a) a Caucasian’s DNA test shows her to be part Native American, part European, and part Hispanic; (b) a geneticist expresses the belief that DNA research will show that all humans are “one”; or (c) an article discussing human migration shows that humans share a common genetic origin.
2. Non-black strengthen: The article suggested that recent advances in genetics or DNA technology could reify or strengthen racial identity, and their subject is implicitly or explicitly non-black. For example, (a) a Caucasian’s DNA test shows that she is not English, but that all her ancestors can be traced to a Slavic tribe, or (b) a Mexican-American feels closer to his relatives when he finds common Sephardic Jewish ancestry.
3. Black blend: The article implied or stated that recent advances in genetics or DNA technology blur or blend racial identity for African Americans. For example, (a) a black man’s DNA test traces his male lineage to Europe, or (b) a geneticist points out how much of the African American population has nonblack ancestry.
4. Black strengthen: The article suggested that recent advances in genetics or DNA technology reify or strengthen racial identity for African Americans. For example, (a) a black woman’s DNA test connects her to the Mende tribe in Africa or (b) a geneticist emphasizes that all human populations stem from a group in sub-Saharan Africa.
6. Both: The article stated or implied both that advances in genetics or DNA technology can blend racial identity and that these advances can strengthen it. For example: (a) the article describes both a black woman’s DNA test that connects her to the Mende tribe in Africa, and a black man’s test that links him to Europe, or (b) two experts debate in the article whether DNA tests blur or reinforce racial distinctions.
6. Neither: The article did not mention these issues, or expressed or implied no view on whether advancements in genetics blend or reinforce racial identity.
The 296 hand-coded articles sorted as follows: 25 in Non-black blend, 6 in Non-black strengthen, 16 in Black blend, 32 in Black strengthen, 26 in Both, and 190 in Neither.
Our theme of medical use of genetic information is more straight-forward, even if the actual medical use of genomics is enormously complex. We hand-coded 238 non-randomly chosen articles from the final database into the following categories:
1. Individualized: The article implied or states that DNA technology or genetic research is making medicine more personalized. For example: (a) a scientist is quoted as saying that DNA technology maps the genome so that a person could get a individualized treatment based on his genetic make-up, (b) a woman gets tested for a breast cancer gene and is given a treatment based on her specific genetic mapping; (c) a doctor is interviewed and explains that DNA technology is leading to “personalized medicine.”
2. Race-based: The article stated or implied that because of DNA technology or genetic research, race remains important or is more important than ever in medicine. For example, (a) a researcher is quoted as saying that blacks are more likely to suffer from colon cancer because of their genetic backgrounds; (b) scientists are quoted as saying that Mexican-Americans in Los Angeles should be more proactive about diabetes check-ups; (c) a article describes many Jewish couples in New York getting tested to see if they carry the gene for Tay Sachs disease; (d) a public health specialist warns that blacks are at a higher risk of heart disease because of an interaction between genetic and environmental factors.
3. Both: The articles stated or suggested that recent advances in genetics or DNA technology can make racial categories both more and less important. For example, (a) an article states that many blacks carry the gene for sickle-cell anemia, but notes that it varies from person to person and that non-blacks could also carry it; (b) the article includes a debate among experts over how much genetics is implicated in disease etiology or treatment.
4. Neither: The article suggested that recent advances in genetics or DNA technology result in neither individualized nor race-based medicine.
5. Silent or Unknown: The article was about something else entirely or was silent on whether genetic research leads to an individualized or race-based medicine.
The 238 hand-coded articles sorted as follows: 58 in Individualized, 35 in Race-based, 4 in Neither, 18 in Both, and 127 in Silent or Unknown.
Results
The ReadMe content analysis program yielded a variety of intriguing, though still preliminary, results.
Social DNA Testing: Once the 296 articles were hand-coded, we used the software package ReadMe to code the remaining 48,092 articles. The results are in Table 1:
Table 1: ReadMe Coding of Final Database on Social DNA Testing
1. Year | 2. Number of articles | 3. Non-black blend | 4. Non-black strengthen | 5. Black blend | 6. Black strengthen | 7. Both | 8. Neither |
1988 | 690 | 7% | 3% | 7% | 6% | 6% | 70% |
1989 | 841 | 6 | 2 | 5 | 5 | 6 | 76 |
1990 | 1058 | 7 | 3 | 6 | 5 | 5 | 75 |
1991 | 1362 | 6 | 2 | 6 | 5 | 5 | 75 |
1992 | 1559 | 6 | 2 | 6 | 4 | 6 | 77 |
1993 | 1904 | 5 | 2 | 5 | 5 | 5 | 78 |
1994 | 2276 | 5 | 2 | 5 | 4 | 5 | 79 |
1995 | 2463 | 5 | 2 | 5 | 5 | 6 | 78 |
1996 | 2770 | 6 | 2 | 6 | 4 | 5 | 77 |
1997 | 3348 | 6 | 2 | 5 | 5 | 6 | 77 |
1998 | 2758 | 5 | 2 | 5 | 4 | 6 | 78 |
1999 | 3297 | 6 | 2 | 6 | 4 | 6 | 75 |
2000 | 4072 | 6 | 2 | 4 | 5 | 6 | 77 |
2001 | 4196 | 6 | 2 | 4 | 7 | 8 | 73 |
2002 | 3390 | 6 | 2 | 3 | 6 | 7 | 75 |
2003 | 3068 | 6 | 2 | 5 | 4 | 5 | 77 |
2004 | 3718 | 6 | 2 | 5 | 4 | 5 | 78 |
2005 | 4767 | 6 | 2 | 5 | 6 | 5 | 77 |
2006 | 4736 | 5 | 2 | 5 | 5 | 4 | 78 |
Several findings are apparent. As column 8 shows, the proportion of articles devoted to the topic of whether social DNA testing strengthens or blurs racial lines is relatively low – never more than a fifth of the total. However, because the number of articles devoted to genomics, medicine, and race is steadily rising (as column 2 shows), readers’ access to articles discussing race and genealogy is growing.
There is preliminary evidence for our expectation that non-blacks and blacks approach the realm of DNA testing differently. Comparing columns 3 and 4 reveals that newspaper coverage of non-blacks who took DNA tests is skewed in favor of linking these tests to blurring or mixing racial identity. Roughly three times as many articles about non-blacks focus on boundary-blurring as focus on boundary-strengthening. Comparing columns 5 and 6, however, shows a different pattern; newspaper coverage of blacks who took DNA tests is roughly equally distributed between discussions of blurring and discussions of reinforcing racial identity. Overall, more articles focus on boundary-blurring (columns 3 + 5) than on boundary strengthening (columns 4 + 6).
In addition, either a disproportionate share of African Americans are interested in this topic, or newspapers are disproportionately interested in African American interest in this topic – we cannot distinguish between those possibilities with these data. For whatever reason, more articles (about ten percent) mention African Americans in relation to DNA testing than discuss all non-blacks, who of course comprise almost nine-tenths of the American population. Reading the articles reinforces this difference in media treatment of black and non-black subjects. Many articles in the “Black strengthen” category employed warm, positive language when depicting blacks able to make a connection to regions of Africa through the use of DNA technology. But articles tended to use a colder, more negative tone when discussing DNA testing’s blurring or confusion of African American lineage. The negative tone was particularly strong in articles reporting on black men whose DNA tests showed a link to Europe rather than to Africa. (The main exception to this generalization was discussion of Henry Louis Gates, quoted earlier as joking about his European ancestry.) Coverage of non-blacks’ use of genealogical DNA tests was nearly always positive in tone.
Note, finally, the striking consistency across time in the proportions of items falling into each category, despite the very great changes since 1988 in genomics research and public awareness of the genomics revolution. This may be an artifact of the mechanized content analysis, which categorizes articles on the basis of phrases and words in our hand-coded set even if the substance of the articles using those phrases or words has changed considerably. We need further analysis, both with and without ReadMe and with different coding schemes, to sort out the issue of change over time.
Medical Use of Genomics Research: For this theme also, we used ReadMe to code the final database, using the hand-coded articles to direct the software. Table 2 shows the results:
Table 2: ReadMe Coding of Final Database on Medical Use of Genomics Research
1. Year | 2. Number of Articles | 3. Individual-ized | 4. Race-based | 5. Both | 6. Neither | 7. Silent |
1988 | 636 | 22% | 14% | 5% | 3% | 56% |
1989 | 787 | 22 | 13 | 6 | 2 | 57 |
1990 | 1004 | 22 | 12 | 5 | 2 | 58 |
1991 | 1308 | 20 | 15 | 5 | 2 | 57 |
1992 | 1505 | 22 | 14 | 4 | 3 | 57 |
1993 | 1850 | 20 | 13 | 6 | 3 | 59 |
1994 | 2222 | 19 | 13 | 6 | 3 | 58 |
1995 | 2409 | 20 | 15 | 6 | 2 | 57 |
1996 | 2716 | 21 | 14 | 6 | 2 | 56 |
1997 | 3294 | 22 | 15 | 5 | 2 | 56 |
1998 | 2704 | 20 | 14 | 6 | 2 | 57 |
1999 | 3243 | 21 | 16 | 6 | 2 | 55 |
2000 | 4018 | 24 | 14 | 6 | 3 | 53 |
2001 | 4142 | 22 | 12 | 6 | 3 | 57 |
2002 | 3336 | 26 | 14 | 6 | 2 | 51 |
2003 | 3014 | 25 | 15 | 5 | 2 | 53 |
2004 | 3664 | 23 | 12 | 5 | 2 | 57 |
2005 | 4713 | 24 | 12 | 5 | 3 | 56 |
2006 | 4682 | 22 | 13 | 6 | 3 | 57 |
Here too, even this preliminary analysis reveals some important patterns. Fewer of the articles – between half and three-fifths—are silent on our topic. The results also suggest a slight upward trend in the proportion of articles devoted to the medical use of genomics research, although any trend there is swamped by the sharp increase in the absolute number of articles written on medicine and genomics. The relative rise and especially the absolute rise both suggest that the American reading public is becoming much more aware of the issue of individualized versus race-based medical research and treatment.
Many more articles – in some years, close to twice as many – focus on the promise of DNA research for personalized medical treatment than focus on the racial basis of disease or treatment. A substantial fraction also discuss both positions, suggesting either that there is a complex interaction between individual and racial etiology or that the journalists are reporting on contention between the two positions. On balance, however, the reporting on medical uses of DNA research seems to be teaching Americans that race matters less in understanding and treating disease and that individual idiosyncrasies matter more.
The proportions in all of these columns are fairly consistent over the two decades. Again, we cannot yet interpret that result: either, given the large number of articles, the proportion devoted to each position is roughly stable despite huge changes in the context and content of genomics or the software is not picking up change over time. This too is grist for a later mill.
Conclusion, Limitations, and Next Steps
Combining the results of the two content analyses leads us to the tentative conclusion that so far, the genomics revolution as reflected through the media is leading Americans to think more in terms of complex group configurations with declining significance of race than in terms of sharpened racial categories with increasing significance of race. More newspaper stories about genealogical discoveries focus on racial boundary-blurring than on racial boundary-strengthening; that discrepancy holds especially for the majority of the population that is not black. More newspaper stories discuss individualized medical research, diagnosis, and treatment than discuss race-based medicine. The key exception, as so often in American society, has to do with African Americans. Their search for their racial or tribal roots, or at least journalists’ attention to African Americans’ search, partially cuts against the tendency toward racial boundary-blurring and the focus on individualized medicine. So far we see little change over time in those findings. But their very consistency in the face of enormous changes in genomics research leads us to suspect that the consistency is a methodological artifact rather than a true finding. For that among other reasons, this paper represents a starting point.
If the tendency toward emphasizing individualization holds up under further scrutiny, it might be part of a broader societal trend toward recognition and celebration of multiracialism [(Williams 2006); (DaCosta 2007); (Hochschild and Weaver 2009)] or even toward post-modern identity fluidity. If the tendency toward distinctive treatment of African Americans’ search for “roots” and perhaps for group-specific medicine holds up, it will reinforce the old American trope of black exceptionalism.
Both results, or any other, could be profoundly important if “the analogy between 20th-century physics and 21st-century biology… continue[s].” Physics, after all, brought us the theory of relativity, the nuclear bomb, and the idea of multiple dimensions; once everyone’s DNA profile is available cheaply, quickly, and accurately, American society will once again be “advancing into the unknown,…for both good and ill.”
Much work remains before we are confident of our current conclusions. Newspaper coverage might be skewed to emphasize particular issue areas, while ignoring other key aspects of the genomics revolution. More analytically, the media do not directly reflect public opinion or elite views and actions; nor do they directly shape public opinion. We need to explicate more precisely just what role the media play in opening up a brand new vista to American readers, and to triangulate media reports with surveys and case studies of crucial actors in this arena.
In addition, while the simple content analyses reported here can reveal central tendencies and proportions, they fail to do justice to the rich and nuanced reporting on how genomics is being developed, used, feared, embraced, and debated. Much remains to be done by simply reading and analyzing many of these 50,000 articles.
We plan also to develop other themes in order to understand the developing politics of genomics. Two topics are especially important. The first is public attitudes toward genetic research. Citizens sometimes embrace scientific innovations, and sometimes fear or mistrust them. Two questions immediately arise: how much confidence does the public have in this new technology, and how do citizens understand and evaluate the normative, political, and policy implications of genomics research? Further questions follow: how rapidly and along what pathways do knowledge of and attitudes toward the genomics revolution diffuse? Do different segments of the population – distinguished by education, ideology or partisanship, race or ethnicity, religion, medical history– hold distinct views or develop views differently? Do citizens seek regulation, or governmental support for medical DNA research, or other policy interventions? How might knowledge of one's own genetic map be associated with a person's understanding of self or identification with groups? Do people associate the genomics revolution with other values, attitudes, or social phenomena?
The second topic is the mirror image of the first: elite attitudes toward genetic research: Roughly the same questions arise: how much confidence do various elites have in this new technology, and how do elites understand and evaluate the normative, political, and policy implications of genomics research? How do these views develop and change over time as elites learn more about the risks and rewards of this new science? A variety of “elites” are relevant here, including geneticists, medical researchers, politicians, policy-makers and regulators, and judges. It may in many cases be appropriate to consider state-level elites as well as people in the federal government.
We conclude by drawing readers’ attention to the two core intuitions driving this research agenda: the perhaps unique importance and complexity of the genomics revolution, and the unusual opportunity to study the growth of a political and ideological system around a brand new issue.
On genomics itself: there are many well-studied cases of the politics of scientific and technological innovation. But we can think of no other case that combines such a high level of knowledge and cognitive sophistication to attain a basic understanding, with so many intensely personal implications in arenas ranging from medicine through law, social networks, family connections, and personal identity. To see this point, compare genomics with two other politically inflected scientific issues. On the one hand, to understand the harnessing of nuclear power requires great knowledge and cognitive sophistication -- but nuclear power does not touch people’s daily lives in intimate ways. On the other hand, environmental degradation may touch people’s daily lives in intimate ways -- but a basic understanding of pollution or global warming does not require great knowledge and cognitive sophistication. Genomics is both obscure and immediate.
Thus existing research on the politics of technological change is useful but incomplete, as is research on the social construction of race, the philosophy of science and medicine, the politics of social policy issues, and the diffusion of innovation. We need new frameworks and evidence to understand the political implications of this new “advance into the unknown.”
Finally, consider the unusual opportunity to study the growth of a political and ideological system around a brand new issue. Here too counterexamples are illuminating. Consider the innovation of stem cell technology. The moment it became feasible and publicly visible, discussion about stem cell research on human embryos became attached to the politics of abortion. Most members of the Democratic Party almost immediately endorsed research with federal funding, while most members of the Republican Party were equally quick to oppose it. Thus within a week of President George Bush’s 2001 speech restricting federal funding for stem cell research, three-quarters of Republicans but only half of Democrats endorsed his decision.[10] In surveys in 2004[11] and 2005,[12] the partisanship of support or opposition remained intact (http://www.pollingreport.com/science.htm). Another example: even Congressional hearings on baseball stars’ alleged use of steroids immediately took on a partisan caste, as Republican members of Congress defended Roger Clemens while Democratic members attacked him.
But despite almost two decades of research and increasing public visibility, genomic research and DNA testing have yet to develop a left-wing or right-wing valence. As we have noted, African Americans are split on whether DNA research will offer new hope for previously-ignored diseases to which blacks are especially prone, or whether it will devolve into a new form of racial profiling. Libertarians and civil rights activists are currently united in their concerns about the privacy of individual medical records, but it is not clear that such an alliance will persist. The search for one’s ancestral roots, one’s dog’s genetic lineage, the names of murder and massacre victims, DNA-compatible marital partners, a cure for diabetes or schizophrenia or breast cancer – none of these and the many other uses of genomics has, yet, a partisan caste.
But the phenomenon is too important and too multi-faceted to remain ideologically and politically innocent for long. Just as real scientists are “advancing into the [biological unknown,” so we political scientists have before us the rare chance to study how a brand new topic gains political and normative valence. Research on public and elite attitudes toward genomics has just begun, largely because attitudes themselves are just beginning to develop. By starting this research now, we can trace the creation of citizens’ and politicians’ views on the medical, racial, partisan, and policy implications of a brand new phenomenon. That chance comes seldom.
Mary Claire King, the discoverer of the BRCA1 and BRCA2 genes, reports that her lab scientists are banging down the door at 7 a.m., they are so eager to see what they will discover that day. We are not quite so excited, or so certain of radical discoveries, but we are not far behind.
Appendix A: Clusters in the Large Database Created by the Expressed Agenda Method
1. First iteration, March 10, 2009:
Proportions Stems
0.06 compani, research, develop, technolog, biotech, year, busi, million, industri, univers
0.059 studi, diseas, heart, health, diabet, research, bodi, risk, fat, diet
0.056 diseas, alzheim,brain,research, gene, studi, patient, caus, cell, year
0.051 cell, research, stem, embryo, human, bush, clone, embryon, presid, life
0.047 test, diseas, genet, screen, famili, babi, disord, medic, children, health
0.045 human, genet, clone, scienc, scientist, research, gene, life, dna, anim
0.044 gene, genet, research, diseas, cell, scientist, human, therapi, mice, caus
0.043 cancer, breast, women, risk, test, studi, year, gene, patient, research
0.042 vaccin, viru, infect, diseas, u, research, scientist, bacteria, human, caus
0.037 genom, human, gene, sequenc, genet, dna, project, scientist, research, map
0.036 drug, patient, compani, health, medic, care, year, fda, cost, trial
0.035 drug, patient, cancer, cell, therapi, blood, treatment, tumor, trial, compani
0.034 cell, cancer, gene, protein, research, tumor, scientist, bodi,dna, caus
0.032 child, babi, parent, children, genet, coupl, fertil, sperm, birth, test
0.031 year, children, time, like, medic, world, book, live, just, health
0.026 genet, insur, test, health, inform, employ, privaci, medic, discrimin, care
0.025 dna, famili, african, histori, test, american, ancestor, ancestri, jefferson, black
0.023 heart, blood, studi, cholesterol, arteri, attack, risk, pressur, diseas, patient
0.023 cell, stem, research, embryon, line, human, tissu, embryo, fund, adult
0.022 center, medic, research, institut, school, hospit, univers, health ,program, care
0.021 anthrax, lab, cow, offici, test, case, diseas, investig, dna, sampl
0.021 compani, abstract, share, stock, drug, patent,agreement, busi, pharmaceut, deal
0.02 cell, stem, embryo, clone, research, human, embryon, egg, creat, tissu
0.019 blood, cell, sickl, marrow, transplant, cord, donor, bone, diseas, anemia
0.019 alcohol, studi, disord, brain, depress, mental, schizophrenia, autism, behavior, twin
0.018 donohu ,skin, dear, caus, iron, pain, hair, doctor, symptom, blood
0.018 human, dna, differ, genet, popul, africa, race, modern, ancestor, 000
0.017 aid, viru, vaccin, hiv, infect, immun, drug, test, trial, strain
0.016 gene, therapi, cell, patient,experi, immun, viru, cystic, fibrosi, lung
0.009 clone, human, embryo, ban, sheep, dolli, egg, reproduct, therapeut, anim
0.009 fibrosi, cystic, lung, autism, children, cf, transplant, child, parent, walk
0.007 chromosom, male, femal, sex, pair, ancestor, evolutionari, evolut, behavior, speci
0.006 coli, outbreak, investig, crime, counti, sampl, fingerprint, offici, forens, fbi
0.004 patent, hood, roanok, ownership, bull, disc, fame, boca, physicist trademark
0.004 embryo, fertil, egg, coupl, vitro, infertile ,implant, reproduct, hugh, adopt
0.004 brain, neuron, rat, nerv, mice, dopamin, parkinson, prairi, rodent, neuroscientist
0.003 weapon, biolog, war, gulf, armi, anthrax, warfar, agent, veteran, iraq
0.003 blood, cord, bank, umbil, marrow, transplant, molli, store, donat, donor
0.003 fragil, homosexu, orient, jacki, epilepsi, bald, cerebr, festiv, mexican, airway
0.002 donor, marrow, ear, gate, african, jone, registri, match, sandi, matthew
0.001 dolli, sheep, lamb, reev, campbel, scotland, silver, molli, beam, wolf
0.001 autism, autist, milk, placenta, coach, china, dame, nile, auction, grandson
0.001 dopamin, pleasur, ferri, reward, lover, hemoglobin, messeng, calm, sandwich, lynch
0.001 donor, marrow, registri, match, gate, monica, neal, suitabl, tammi, raiser
0.001 gallon, mayor, gail, ounc, drink, tangl, calcium, underwrit, golfer, consumpt
NOTE: We directed the software to eliminate shaded items in the next round of clustering, on the grounds that the articles captured by that set of words would be irrelevant to our current research questions. That process of elimination led to the final iteration, below.
2. Final iteration, March 14, 2009:
Proportions Stems
0.061 heart, diseas, patient, blood, caus, doctor, bodi, medic, year, pain
0.054 diseas, alzheim, brain, gene, research, studi, caus, ag, cell, patient
0.053 research, medic, center, health, univers, institut, year, hospit, school, care
0.049 alcohol, studi, children, autism, genet, famili, disord, time, like ,health
0.046 cancer, breast, tumor, patient, cell, women, research, studi, treatment, drug
0.042 clone, human, cell, research, embryo, anim, scientist, stem, genet, creat
0.04 cell, research, human, gene, scientist, protein, mice, organ, bacteria, anim
0.04 gene, human, genom, dna, genet, scientist, sequenc, chromosom, research, differ
0.04 genom, human, genet, gene, project, sequenc, compani, dna, research, patent
0.04 studi, heart, caus, diseas, journal, research, gene, blood, risk, infect
0.038 gene, therapi, cell, patient, trial, experi, treatment, research, immun, blood
0.037 cell, stem, research, embryo, embryon, human, clone, line, scientist, fund
0.037 vaccin, viru, aid, infect, hiv, immun, cell, research, antibodi, virus
0.036 drug, compani, patient, approv, fda, market, trial, product, test, cost
0.035 test, babi, children, screen, child, parent, newborn, disord, birth, hospit
0.035 diet, fat, food, studi, cholesterol, heart, eat, diabet, weight, obes
0.029 babi, women, test, pregnanc, embryo, fertil, birth, egg, child, coupl
0.029 blood, cell, sickl, marrow, transplant, donor, bone, cord, diseas, match
0.028 vaccin, viru, flu, infect, bird, mosquito, strain, diseas, world, malaria
0.028 gene, diseas, genet, test, fibrosi, cystic, famili, disord, children, defect
0.028 insur, health, genet, care, test, inform, medic, employ, privaci, compani
0.027 cell, cancer, gene, protein, tumor, dna, chromosom, scientist, normal, mutat
0.026 human, dna, year, ancestor, modern, africa, 000, speci, popul, evolut
0.023 cancer, breast, women, risk, test, amili, mutat, genet, histori, ovarian
0.02 mice, brain, mous, alzheim protein, muscl, anim, nerv, dystrophi, muscular
0.016 comput, space, devic, war, technolog, nuclear, remain, air, centuri, engin
0.015 heart, diabet, risk, black, cholesterol, pressur, exercis, high, women, american
0.013 dna, african, famili, ancestri, ancestor, trace, histori, black, sampl, descend
0.011 fibrosi, cystic, lung, cf, transplant, mucu, foundat, breath, walk, digest
0.007 jefferson, father, african, descend, ancestri, famili, dna, patern, adopt, thoma
0.007 abort, evolut intellig, god, creation, moral, religion, religi, christian, steril
0.006 insur, discrimin, employ, deni, coverag, prohibit, counselor, hire, predisposit, jordan
0.002 kimberli, rebecca, reunion, arrest, jessica, paula, juri, wright, custodi, vista
0.001 roch, needi, softbal, transact, herb, airway, stanley, essay, buy, jeffrey
References
Burchard, Esteban et al. 2003. "The Importance of Race and Ethnic Background in Biomedical Research and Practice." New England Jounal of Medicine. 348 (12): 1170-75.
Cavalli-Sforza, L. Luca, Paolo Menozzi, and Alberto Piazza. 1994. The History and Geography of Human Genes. Princeton NJ: Princeton University Press.
Cooper, R.S. et al. 2003. " Race and Genomics." New England Jounal of Medicine. 348 (12): 1165-70
DaCosta, Kimberly. 2007. "Making Multiracials: State, Family, and Market in the Redrawing of the Color Line." in ed. Stanford CA: Stanford University Press,
Duster, Troy. 2005. "Race and Reification in Science." Science. February 18, 1050-51.
Economist. 2007. "Biology's Big Bang." June 16, 13.
Gates Jr., Henry Louis. 2006. "My Yiddishe Mama." Wall Street Journal. New York. February 1, xx.
Grimmer, Justin. forthcoming 2009. A Bayesian Hierarchical Topic Model for Political Texts: Measuring Expressed Agendas in Senate Press Releases. Cambridge MA: Harvard University, Government Department.
Harmon, Amy. 2006. "Love You, K2a2a, Whoever You Are." New York Times. New York. January 22, sec. 4, pp. 1, 18.
Hitt, Jack. 2005. "Mighty White of You: Racial Preferences Color America's Oldest Skulls and Bones." Harper's. July, 39-55.
Hochschild, Jennifer and Vesla Weaver. 2009. Multiracialism, Feedback Effects, and the American Racial Order. Cambridge MA: Harvard University, Department of Government
Hopkins, Daniel and Gary King. 2008. A Method of Automated Nonparametric Content Analysis for Social Science. Cambridge MA: Harvard University, Department of Government.
Kalb, Claudia. 2006. "In Our Blood." Newsweek. February 6, 47-55.
Kohn, Marek. 2006. "The Racist Undercurrent in the Tide of Genetic Research." Guardian. London. January 17, 26.
Lei, Hsien-Hsien. 2007. DNA-Supported Ancestral Tourism. DNA and Genealogy. (http://www.eyeondna.com/2007/06/11/dna-supported-ancestral-tourism/).
Lomax, John 2005. "Who’s Your Daddy? Track Your True Identity along a DNA Trail Left Behind by Your Ancestors." Houston Press. Houston TX. April 14, xx.
Nathan, David. 2007. The Cancer Treatment Revolution. Hoboken NJ: John Wiley & Sons.
Science. 2007. "Closing the Net on Common Disease Genes." May 11, 820-22.
Smith, Stephen. 2009. "MGH to Use Genetics to Personalize Cancer Care." Boston Globe. Boston MA. A1, A7.
Wade, Nicholas. 2005. "Researchers Say Intelligence and Diseases May Be Linked in Ashkenazic Genes." New York Times. New York. June 3, A21.
Williams, Kim. 2006. Mark One or More: Civil Rights in Multiracial America. Ann Arbor MI: University of Michigan Press.
Willing, Richard. 2006. "DNA Rewrites History for African-Americans." USA Today. February 1, xx.
WordNet® 3.0. n.d. Genomics.
Yancy, Clyde. 2007. "The Association of Black Cardiologists Responds to "Race In A Bottle:" A Misguided Passion." Scientific American.
[1] “the branch of genetics that studies organisms in terms of their genomes (their full DNA sequences)” according to the simplest definition (WordNet® 3.0 n.d)
[2] A good example of an exchange of views on this issue is in (Cooper 2003) and the reply in (Burchard 2003)
3] We have tentative approval to include 22 items on the societal implications of DNA on the 2010 General Social Survey, and we are applying for support to replicate them on the 2012 GSS. A few poll items on genomics and its implications are beginning to emerge.
[4] The actual search terms were: “(DNA or genetic! or genom!) w/10 (medic! or illness! or disease!) and not (rape! or murder! or police! or defendant! or dog!).” We requested the exclusions in order to avoid both the context of the criminal justice system and criminal forensics, which lies outside the scope of this paper, and the frivolous but surprisingly popular pastime of tracing pets’ genetic heritage.
[5] The actual search terms were “(DNA or genetic! or genomic!) w/10 (race or racial! or ethnic!) and not ("race to" or rape! or murder! or police! or defendant! or dog!).”
[6] For this search, the search terms were “(DNA or genetic! or genomic!) w/10 (geneal! or root! or ancest! or nationalit!) and not (rape! or murder! or police! or defendant! or dog!).” Word stems of keywords related to the use of DNA or genetic information in the course of criminal forensics were removed from the search parameters.
[7] In the revision of this paper, we will update the databases through 2008 and perhaps into 2009.
[8] The articles were analyzed using the statistical program R and software code written by Justin Grimmer. The software will soon be available as a downloadable R package. Grimmer (2009) provides full details of this method.
[9] The program used here was ReadMe, a package for the statistical program R. ReadMe is available for free download at http://gking.harvard.edu/readme.
[10] Ipsos-Reid Poll, Aug. 10-12, 2001, N= 1000 adults nationwide
[11] University of Pennsylvania National Annenberg Election Survey. July 30-Aug. 5, 2004. N=1,345 adults nationwide.
[12] CBS News Poll. July 13-14, 2005. N=632 adults nationwide
