Prediction as the Cure
Gazing into the human body’s crystal ball

Vol. 10 No. 3
December 2007/January 2008

Return to Contents

Prediction as the Cure
Gazing into the human body's crystal ball

The "engineers" in bioengineering

“I find it interesting that it’s incredibly difficult to define health without talking about a default position—you know you have health because you’re not sick.”

“There’s significant new ethical ground to be broken here. . . . We’re collecting a large amount of personal information. How do you protect it, who stores it, how do you store it, who has access to it?”

Defending basic science

Looking South Exploring Southern Spaces
The new terrain of Emory's multi-media scholarly journal

The Near Past
Tone and tension in writing about the modern South

Stories from the Frontlines
Or, How to survive co-authoring a book (with your spouse


Is health simply the absence of illness? At what point does a person move from the realm of health to that of un-health, and can we foretell, with any practical precision, when they will cross the boundary? Michelle Lampl
has given the matter a lot of thought. “I find it interesting that it’s incredibly difficult to define ‘health’ without talking about a default position—you know you have health because you’re not sick. It’s a big challenge for science to be able to begin to discuss a concept of normality without it being the mean or the average,” says the Samuel C. Dobbs Professor of Anthropology, who is also a physician. For Lampl the anthropologist, it’s a beguiling intellectual exercise. And the possibility that science may be able to predict in some detail a person’s comprehensive health destiny tantalizes Lampl the physician.

Lampl is associate director of the Predictive Health Initiative (PHI), a new collaborative undertaking with Georgia Tech that “focuses on maintaining health rather than treating disease.” Applying genetics, proteomics, and computational biology, PHI researchers aspire to identify chemical markers (or “biomarkers”) that augur physiological aberrations long before they manifest as illness, then intervene in time to change the natural course of a disorder. Short-circuiting disease processes may be as simple as changes to diet and exercise, drugs, or, farther along, gene therapy or other advanced techniques.

But the path to this ambitious goal is strewn with potholes. Privacy advocates worry about the legal and ethical implications of large-scale, centralized accumulation of personal information. Substantive changes in how health care is delivered threaten to render archaic state and federal laws that regulate health care, health insurance, and determination of court awards in some types of cases. Even epidemiology, the science upon which much of predictive health is founded, is subject to notable limitations when it comes to determining disease cause and effect.

Predictive medicine in and of itself is nothing new: cholesterol is a long-standing indicator of future cardiovascular disease; high blood pressure portends a variety of dangerous conditions. Genetic markers can be highly prognostic of future disease. The hunt is on for genetic markers for cardiovascular disease, colon cancer, and diabetes. But what is new is the focused, systematic attempt to move predictive health—and therefore preventive health—into mainstream medicine. The attendant benefits of such an epic shift in mindset could reach beyond decreased individual suffering to include lower medical and insurance costs, and fewer sick days.

Discovery and definition

The most visible facet of the PHI is the new Center for Health Discovery and Well Being in midtown Atlanta, which will soon begin enrolling its first cohort of healthy adults—hence the importance of deciding who fits the bill. The PHI deliberately defined “health” loosely, anticipating that future tabulations of data for thousands of people over many years will lead to a more focused characterization, says Kenneth Brigham, associate vice president and interim director of the PHI, and vice chair for research, Department of Medicine. “We’ve wrestled with this question [of who to enroll],” says Brigham, a pulmonologist. “We’re not going to rule out people with well-controlled chronic disease like hypertension or diabetes and who are being followed by their doctor, but we won’t treat them either. We would exclude people whose illness is of great enough severity that they couldn’t participate in the program, such as those with recent series of heart attacks or who are in current therapy for cancer.” The PHI plans to populate the first cohort with Emory employees, at an initial cost of $5,000-$7,000 per person (visit for enrollment details). Emory has committed about $10 dollars to the PHI, while Georgia Tech will provide further funds and faculty appointments. Additional money will come from philanthropic organizations.

Each participant will be matched with a health partner—someone well trained for the task, though not a physician—to shepherd them through a comprehensive work-up, develop a personalized health plan, and monitor progress. They’ll complete a series of questionnaires to evaluate physical and emotional health and even spirituality. Weight, height, skin conductivity, bone density, and body fat will all be recorded. An ultrasound will be used to evaluate the thickness of the carotid artery (a surrogate marker for atherosclerosis). Other assessments include blood vessel reactivity and central blood pressure; a modified fitness test; and blood work for lipids, glucose levels, and another four dozen compounds that might provide clues to future health. The data will be stored and available to researchers for mining and analysis. One of the planned studies will compare center participants to nonparticipants to see how each group fares down the road.

The PHI will also include sponsored research. Roberd Bostick, a professor of epidemiology in the School of Public Health as well as of hematology and oncology at the Winship Cancer Institute, is studying biomarkers that predict who is at highest risk for colon cancer. “Rates of cardiovascular disease are declining very sharply, and one of the primary reasons is biological markers such as cholesterol, but we have no such markers for risk of cancer, and cancer rates have changed very little,” Bostick explains. “We wanted to do something analogous for colon cancer and prostate cancer.”

Kathy Parker, Edith F. Honeycutt Professor of Nursing and co-director of the Emory Sleep Center, is seeking to develop predictive genetic models to determine which opioid painkillers, and at what doses, provide the optimal combination of analgesia with the smallest impact on sleep among cancer patients. “Sleep,” says Parker, “is a fundamental aspect of health, and poor quality or insufficient sleep is strongly associated with poor health outcomes.”

Much of the scientific grounding of the PHI rests in epidemiology, which attempts to identify the causes of chronic common diseases. Lately, however, some are beginning to question whether medicine relies too much on epidemiology to draw conclusions about what makes us sick or keeps us healthy. A September 16 New York Times Magazine article noted that epidemiologists gather data from large groups of individuals in so-called observational studies and tease out how particular factors influence health using statistical analyses. But, says the Times article, no matter how well designed and how many tens of thousands of subjects they might include, their fundamental limitation is that while they can distinguish associations between two events, they cannot inherently determine that one event causes another. As a result, observational studies provide what a defense attorney would call circumstantial evidence. The article goes on to say that “because these studies can generate an enormous number of speculations about the causes or prevention of chronic diseases, they provide the fodder for much of the health news that appears in the media.” Those speculations “play a significant role in generating public-health recommendations”—as well as some key emphases of the Center for Health Discovery and Well Being.

What is private?

At some point, the center will conduct genetic analyses on the participants’ blood samples. The prospect of the methodical collection of human genes disturbs many people at a gut level. After all, a health insurer that learns an applicant has a genetic predisposition for a condition that will one day require costly treatment would likely deny or severely limit coverage, or charge huge premiums. “If you have a regime where you’re looking at the genetic makeup of individuals and possible diseases they might have, then you have to consider the possibility that genetic information might reach employers or insurance companies,” says David Partlett, dean of the Emory law school. “On the other hand, if you know someone has a genetic marker for seizure disorders or alcoholism, would you want to give that person the captaincy of an oil supertanker, which was the case with the Exxon Valdez?”

Ani B. Satz, an associate professor with joint appointments in the law school and School of Public Health, doesn’t believe that genetic information deserves its suspicious repute, at least when it comes to privacy and confidentiality concerns. “Genetic information is simply one form of medical information,” she says. “It raises the same concerns as other nongenetic, predictive health information that is shared. The view that there is something different about genetic information is called genetic essentialism, which, though once the dominant view, is now largely rejected. If we need heightened privacy protections, we need them across the board, not specifically with regard to genetic information.” Though cholesterol and blood pressure aren’t as arcane, they are nonetheless predictive of heart disease. Nongenetic information about contagious disease and environmental toxins may also be shared, contends Satz. HIV infection, for example, may be passed from mother to child. Individuals who live near a toxic waste site may be exposed to carcinogens or other chemicals (dioxin, for example) that might harm them and their progeny.

Control of health information is just one problem. If predictive health gains a significant foothold and leads to material changes in our health culture, it could generate other knotty legal and political issues. An increase in life expectancy—a long-term outcome that predictive health advocates expect—would strain government benefits such as social security and play havoc with personal injury awards. “Tort law is largely written in such a way as to focus on disease and not health,” said Partlett, because courts often award compensation in injury cases based upon the victim’s reasonable life expectancy and expected working life. Longer life spans would therefore dictate greater awards. “We complain about tort damages now being excessive, but they’re likely to increase,” he adds. On the other hand, medical advances, such as a therapy that reverses limb paralysis, could offset the financial impact of longevity.

Any compelling change in our health care system will necessitate a re-evaluation of its legal underpinnings. That will be one of the charges of the Center for Health Law, Policy, and Ethics (CHLPE). Satz, who is spearheading the project, hopes it will attract preeminent scholars in all three subdisciplines. CHLPE would consider, among other things, the regulation of health care delivery, which profoundly affects how a government interacts with its citizens, says Partlett: “How will you fund health care, how should Medicare be changed, how do you provide service to the needy? Lawyers are asked to solve those problems.”

It’s far too early to do anything but speculate about the enduring influence of predictive health. To reach a critical mass will take “a true paradigm shift on a huge cultural level,” and education is the key, says Lampl. “You can’t just say, OK, you’re going to change the way you think about health, but I think people want to care for themselves, and they’re looking for information about how to do that.” Lampl would like to see an integration of comprehensive health information into school curricula. “There’s no way that people will begin to think about health care in ways other than the emergency room or visiting the doctor when they get sick unless we provide them with alternatives to discuss early in the educational process.”—S.F.