Duke University Alumni Magazine

Listening to our Elders

With the National Long Term Care Survey and other projects, Duke demographers rack up statistics on aging and health, and make some surprising discoveries about getting older and better.

wo years ago in Tucson, Arizona, Duke demographic researcher Larry Corder listened in on a phone conversation. A field representative with the U.S. Census Bureau was asking a sixty-six-year-old woman about her ability to perform household tasks: Could she dress herself? Take a shower? Were stairs a problem? What about shopping for groceries? After only a few of these questions, the woman became impatient. "Look. Just put down that I can do anything," the woman told the caller. Fact is, she was late for a tennis date. Never mind that the temperature outside was 110 degrees Fahrenheit.

Later, Corder went with another representative to see a woman who was ninety-two. Though the woman had been swimming that morning, Corder found her on the patio-smoking a cigarette. "Four a day keeps my weight down," the woman explained to the shocked visitors. Then she told the story of how she was working at the New York Stock Exchange on October 29, 1929, the day of the crash. "You can quote me, Sonny," she said. "That was a bad day." Corder happens to find old folks interesting. Maybe it has something to do with the fact that he's getting older-just like the rest of us. But it's also his job. Corder is a researcher at Duke's Center for Demographic Studies. Led by director Kenneth Manton A.M. '71, Ph.D. '74, the center has been a world leader and an indefatigable generator of data in the field of aging and health.

At the core of the center's work is the National Long Term Care Survey, a longitudinal health assessment of Americans sixty-five and older. So large is the survey-including about 42,000 individuals-that the center hires the Census Bureau to conduct it. If you're among those chosen for the sample, the folks at the center will find you. "The only place we don't go is prisons," explains Corder.

The people we call "old" defy stereotypes. It's no longer any big shock to read about triatheletes in their seventies, stockbrokers in their eighties. Just by looking around-at our parents, our neighbors, maybe even ourselves-we have a strong feeling that those who are living longer are also living better. But this optimistic attitude toward aging hasn't always prevailed. And until the center's researchers began studying the issue, it couldn't be proved.

Many have wondered what happens inside the center's little white house on Campus Drive. Occasionally, someone will be curious enough to pay a visit, like the minister who wanted advice on handling his diverse congregation, or the two visitors who wanted to know the latest number of Native Americans on U.S. reservations. The researchers are willing to help the occasional browser of statistics, but it helps to know the center's focus. Over its twenty-eight-year history, the center's thirty full-time and twenty associate researchers have combined to produce more than 700 books and papers, with no sign of slowing down. Manton's own dry-erase board is covered with more than a dozen ongoing projects, including five for which he is principal investigator.

The Center for Demographic Studies took an interdisciplinary approach before interdisciplinary was cool. Manton was a mere post-doctoral student when he caught the attention of the center's late director, George Myers, who gave him the task of analyzing data sets for one of his projects. It was a portentous move. Besides taking over the directorship from Myers in 1997, Manton would go on to be principal investigator on no fewer than seventy-seven grants and contracts. Later, when it became clear the center needed someone with actuarial experience, Eric Stallard '71, a computer programmer by training, became certified in that area. "'Interdisciplinary' is the buzzword now, but back then we needed it to answer the questions," Stallard says. All of the researchers seem like Renaissance men and women in their own rights, but associate professor Beth Corder (who is married to Larry) has all the degrees, too-in microbiology, pharmacy, statistics, and epidemiology, not to mention a post-doctorate in neurology.

Photo: Les Todd

Myers had come to Duke's sociology department in 1968 from the World Health Organization. "One of the first things he recognized was that demographic research was going to be highly computer-driven," says Stallard. "Also, he realized that aging would become a major issue in the next century." Through his government connections, Myers was able to get his hands on an early release of the 1969 statistics from the National Center for Health Statistics. Essentially, it gave the center the largest mortality data file in the country. That turned out to be the easy part. The next step was coming up with the methods and models to analyze the data, which totaled two million records each year-"not something you can just download into Excel," says Stallard.

The models they created to analyze the mortality data turned out to be versions of Brownian, or seemingly random, motion, similar to the model for pricing stock-options. "These are applied mathematics," explains Manton. "It's nonlinear mathematics, chaotic systems, and nonlinear dynamics that are interesting to me." Not everyone can claim to be one of the world's leading experts in mathematical population models, but Manton can. By the early 1980s, the researchers knew that if they were going to study longevity, they needed to generate more data. The center had been hired by the federal government to analyze the results of a survey conducted in 1982 and 1984 by the Health Care Financing Administration (HCFA). The National Long Term Care Survey, as it was called, included a large, national sampling of adults sixty-five and older. What's more, the survey was designed to follow up on previous survey subjects, replacing those who had died with those who had recently turned sixty-five. It was a brilliant idea. Then the government decided to drop it.

That's when the center stepped in. Before coming to the center in 1987, Larry Corder had worked at HCFA as a statistician developing the 1982 survey. "Ken Manton and I wrote a letter saying, 'If you don't want to do this, we think we will,' " Corder recalls. Manton wrote the grant application asking the National Institute on Aging for just under $5 million to conduct the survey in 1989, and he repeated the process for surveys in 1994 and 1999.

Few questions have changed. "By keeping things small, you can keep them efficient," Manton explains. The survey currently runs up to seventy-five minutes and takes place in people's homes. It has a phenomenal 95 percent response rate, perhaps because the elderly are delighted that someone would take an interest in their health. Questions cover health function, care providers, and finances. Specific health histories of subjects can be accessed through Medicare records. Due to the sensitive nature of its data, the center is one of the securest computer operations on campus. "We've got a double firewall," Manton stresses, meaning that even if a hacker gets into the center's computer system, there's no room for the hacker to move around.

A goodly chunk of the individuals have been part of the survey from 1982 through 1999, a seventeen-year period that happens to represent the average remaining life expectancy for those who reach sixty-five. Exercise, smoking, drinking, and nutrition are just a few of the factors studied. The center's research, among other things, has put into focus the way most old people die.

Photo: Les Todd

"You think of someone getting a disease and dying. That's not what happens to elderly people," Stallard says. "The reason old people die is they contract medical illnesses that lead to a sequence of events that leads to a terminal decline." A person with diabetes, for example, may more easily develop heart disease, which leads to a fatal heart attack. It helps to think of death as a process, a process that may have started before birth.


Once upon a time, there were rocking chairs and porches. Then putting greens were the big draw. Fast-forward to the present, when-who knew?-classroom desks are in vogue for those who have retired their careers but not their minds.

"When you retire at sixty, you can expect to have twenty-five more years of active life. That's too long to just play golf or bridge," says Sara Craven, director of the Duke Institute for Learning in Retirement (DILR). "Now we're moving into this phase where people are doing other things." DILR was founded in 1977, when, as Craven puts it, "it was quite a radical thought that older people not only could go back to school, but that they wanted to go back to school." Forty-two people signed up, and since then the number has grown 10 percent each year. Today, more than 700 people are members of DILR, which is just one of 260 learning institutes around the country.

"Many people say to me, 'Do you know that DILR saved my life?' " Craven says. Losing a spouse can mean losing the link to a social life. DILR provides a network of friends based on common interests, not just age. In some ways, she adds, the program can be considered "a secular church." Herb Halbrecht, seventy-three, owes his decision to retire in Durham largely to DILR. He wanted to be near a golf course and a top-ranked medical center, but learning opportunities were also essential to the equation.

"The role of a university is more than churning out research or educating young people," says Halbrecht. A former recruiter for telecommunications companies, he takes five courses a year at DILR and has taught two himself: "The History of Zionism" and "Joseph Campbell's Mythology." He has also played a role in bringing WebTV to a Methodist retirement home in Durham, and is currently working to computerize schools for the Haliwa-Saponi tribe of Native Americans.

Classes are held at the Bishop's House on East Campus and at satellite locations such as The Forest at Duke. Half are taught by members. They teach what they did in working life, or bring in guest speakers. Some find a hobby area, study it well for a couple of years, then teach an introductory course. Faculty, graduate students, and local experts also fill teaching spots.

None do it for the money. DILR instructors earn $300 for twelve weeks of classes, while members teach for free. A member can take up to five classes a term for the annual fee of $125. That's less than one percent of the cost of a Duke undergraduate education. The university could learn a lot from DILR, says Craven. "The students bring so much to class. What a cliché that youth is wasted on the young, but there are ways in which that cliché is true. The humanities, for example, are just tailor-made for older people who on some level ask, 'What is this all about?' "

-Eric Larson

Once it became clear in the early 1980s that America's elderly were living longer, people began to worry about their health. "There were a number of papers that said we were going to have an elderly society that is very debilitated," says Stallard. Longer lives were bound to deplete Social Security and create a doddering populace, right? In January 1983, the bipartisan National Commission on Social Security (a.k.a. the Greenspan Commission, thanks to its chairman Alan Greenspan) reported the mass of evidence that longevity was increasing. Manton was one of those who testified to the point. Unfortunately, the commission lacked conclusive future-predicting data on the elderly's health, so its report to Congress recommended an increase in the full-benefit retirement age for some, coupled with an increase in taxes for most. Hindsight judges that decision. "If the Long Term Care Survey had been in place, it would have increased the retirement age, but we would have experienced lower taxes," Larry Corder postulates. It just so happened that the 1989 survey was going to defy an assumption of the Greenspan Commission. Instead of the elderly's collective health worsening with greater life expectancy, their health was improving. "It called into question the idea that we had a rising pandemic of mental and physical disorders," says Stallard.

The center wasn't the only clarion trumpeting a decline in chronic disability. Economist Bob Fogel at the University of Chicago researched the health records of Union Civil War veterans born between the years 1825 and 1844. Comparing those soldiers' records with those of veterans of subsequent wars, Fogel established a .6 percent decline in chronic disability each year between 1910 and 1985. Then in March 1997, the center released a paper based on its 1994 survey called "Chronic Disease Trends in the Elderly United States Population, 1982-1994," showing how from 1989 to 1994 morbidity had declined a whopping 1.5 percent each year. "The resistance in the research community was blown away. It took us seventeen years to essentially reverse what the scientific community was thinking about the health and functionality of older persons," says Stallard.

The 1999 survey shows the trend continuing. Fewer smokers, new medical procedures and drugs, and improved diet and food preparation have all contributed to better health for the elderly. The center's detailed findings, scheduled to be announced in a major scientific journal, may have phenomenal implications for national policy. "The National Long Term Care Survey conducted by the Center for Demographic Studies, headed by Ken Manton, is perhaps the only rigorous study that is continuing to provide the economic data necessary for policymakers to make sound decisions for Medicare reform," says Arthur Ullian, chair of the Task Force on Science, Health Care, and the Economy.

Ullian credits the survey with encouraging Congress to double the budget of the National Institutes of Health by 2003. "Manton's work on tracking the rate of declining disability (and conversely, the increase in active life expectancy) will, I believe, become the basis for a Medicare reform that encourages the development and use of new therapies and cures. These, in turn, will lead to even further disability decline, keeping Medicare solvent through 2070, the end of the baby-boom era." "If you can reduce the incidence of disability 1.5 percent a year through 2025 and 2030, then very long solvency can be maintained for Social Security and Medicare," predicts Manton. "You might only need some reserves to get over the hump. This is very controversial stuff, but there's beginning to be more and more data to suggest it."

As it turns out, people showing the signs of age can be grouped into six categories, from those just having trouble climbing stairs (Group One) to those needing full-time care for dementia, stroke, or other disability (Group Six). With each of the past three surveys, the mildest group-those having trouble with stairs only-has become more and more representative of the whole, while the segments with the most serious conditions are becoming consistently less represented. In other words, though the elderly are still manifesting the debilitating effects of aging, they're showing them in less drastic forms.

"Every time we do the survey, the sixty-five- to sixty-nine-year-olds have been healthier than the same group five years before," Larry Corder says. The responses of this particular "young old" group are especially noteworthy. Being surveyed for the first time, they are less likely to report good health simply to please those asking the questions. The upshot is that "people perceive themselves as healthier than those who came before them perceived themselves."

It's easy enough to propose reasons. Beth Corder points to medical advances that mitigate high blood pressure, stroke, and heart disease, as well as a greater willingness in the medical community to treat older people. "Twenty years ago, if you were old and at the Mayo Clinic, you may not have been diagnosed with diabetes, for instance," she says. Doctors would tell the person to go home and rest-in other words, to die. "The idea that you can be old and treated is actually very new. It's a whole paradigm shift."

Thanks to this new paradigm, the fastest- growing segment of America includes those over eighty-five, a population that is doubling every twenty years. (In the long-term care survey alone, 1,000 of the participants are over ninety-five.) Of course, with increased age comes a greater risk of Alzheimer's disease. Between two and three million people have been diagnosed with Alzheimer's, which has a ten-year survivability on average and potentially huge consequences for the costs of long-term care.

Fortunately, the likelihood that neurological disorders can be treated with pharmaceuticals has companies racing to develop drugs regulating the lipoprotein apoE. Medical researchers know that different forms of the lipoprotein are associated with different risks of getting the disease. What isn't known is the prevalence of the various forms in the population.

That's why, as part of the 1999 survey, the center's researchers decided to gather 1,000 blood and mouth-swab samples from subjects. "When detailed information is known, it might tell you when to do preventive medicine for a certain subset of the population," according to Beth Corder, who was the lead author on the landmark 1993 Science article proving the significance of apoE in Alzheimer's. The center is already making efforts to safeguard the DNA samples while making them available to other research institutions.

Hope for dementia treatment now makes those early 1980s predictions look even more short-sighted and cynical than before. The long-term care picture is becoming far more varied and inviting than people ever imagined. "The traditional model of a nursing home where lots of old people go to die is history," says Larry Corder. Nursing homes will still exist in the future, but they will increasingly have rehabilitation as a strong component.

It doesn't hurt that today's elderly are better off financially than their parents were, and the market has responded-a residential care center like The Forest at Duke is just one option unimaginable thirty years ago. For those who lack solid finances, the picture is not so gloomy either. "We see no evidence that people abandon their relatives. In fact, we see that the vast majority of caregiving is done by relatives for no pay."

There are still uncertainties. The sheer numbers within the baby boomer population will mean a building frenzy in the number of long-term care facilities, just as the building projects of obstetrics wings, elementary schools, and colleges have followed boomers through their lives. In addition, the demographic center expects to see a one-year gain in life expectancy, on average, for each of the next seven decades. That means that in seventy years, the average person will receive a Social Security check seven years longer than the seniors of today. Assuming a prescription drug benefit is passed by the next Congress, what will the total price tag be down the road, and can we afford it?

Questions concerning longevity also remain, such as the degree to which we are born with our future already written. To what extent do our genes determine health? "Longevity is only partly inheritable," cautions Larry Corder, pointing to the large body of twin research that can help separate the factors of genes and environment. However, even twin research has fallen under scrutiny recently. David Barker of the University of Southampton in England argues that uterine environments for twins can produce strong differences in long-term health. Because of the role a mother's diet might play, demographers are trying to gather prenatal care data for today's elderly. Manton points to one possible source of such data: Christian missionary hospitals in China, which in the early 1900s had a policy of weighing pregnant women and measuring their bellies throughout their term of pregnancy.

Manton for one isn't so sure that we've seen the limit on longevity. The current threshold for age seems to be 129, with mitochondrial DNA being the weak link in the longevity chain. But our nuclear DNA seems much more resilient, theoretically lasting us up to 300 years. Reaching such biblical ages might sound ludicrous, but science has a way of breaking down long-held beliefs. "There are a lot of old saws getting turned over," Manton says. "Everyone always thought that the brain didn't grow new cells, for instance. But we know now it does, and we've seen evidence for brain stem-cells."

Manton and his team are also bringing brain-power to bear on economic issues, such as the long-term impact of research dollars on health-care savings and the differing rates of chronic disability for minorities and women. Manton's paper on women's health this year won an award from the Partnership for Women's Health at Columbia University by showing a link between women's diminished finances in old age and their increased chance of disability due to cardiovascular disease. Manton found that for a given level of disability, women are more likely than men to be sent to a nursing home. Wider use of hormone-replacement therapy, which can bolster a woman's defense against disease, may be one step toward rectifying the situation.

The Center for Demographics Studies has traveled far in answering questions of disease and aging. But how has such a productive research center remained a sometimes unnoticed worker for so long? "If you can go faster than the speed of light, you can get in Nature very easily," says Manton. "Those types of things are very sexy. You're upsetting Einstein. But if you document that there has been a decline in chronic disability, that's not." That may change. Give the editors of the major scientific journals time-enough of it-and don't be surprised if they become more interested in aging, too.

Larson '93, a regular contributor to the magazine, turned thirty in November and is getting older by the day.


Max Woodbury, professor emeritus: "I really believe it's 'use it or lose it.' So I think all the time."
Photo: Les Todd

Much of the work at the Center for Demographic Studies has to do with complex mathematical models, such technical concepts as "trajectories of aging," "grade of membership," and "pattern recognition." One creative force behind much of the math is Max Woodbury. At eighty-three, Woodbury is one of the eldest professors emeriti still taking an active role in research at the university. He's also an example of the arbitrary limits we sometimes place on age.

Woodbury began his academic life teaching mathematics at the University of Michigan. That was in 1948. He had already spent the war in North Africa in the weather service, helping such military operations strategists as Thornton Wilder plan the U.S. invasion of Italy. Woodbury entered mathematics when computers were just beginning to gain a research following. On November 4, 1952, while teaching at the University of Pennsylvania, he demonstrated for the nation an amazing use of the new technology. Feeding in the evening's earliest election returns as data, he used a program he had written for the UNIVAC I computer to attempt to predict the winner of the Dwight Eisenhower-Adlai Stevenson match-up.

Woodbury had used state records as well as Associated Press teletype returns from past elections to write his program. By 8:30 p.m., the computer was predicting for Walter Cronkite and the rest of the country a 100-to-1 certainty that Eisenhower would win the election, even though Stevenson was leading in votes. As it turned out, the computer's predictions were only one percent off the actual electoral vote count and 3 percent off the popular vote. By 1956, all the TV networks were using computers in their election coverage. Meanwhile, Woodbury developed an interest in medicine. When he came to Duke in 1966, he was asked to teach in both the math department and the medical center. After five years, he was instructing in the computer science department as well.

Some have tried to squeeze together Woodbury's interests and call him a biomathematician, but he prefers the label "scientist of ideas." He observed early that scientists were sometimes boxed in by their disciplines, but a science of ideas that transcended fields was not an accepted concept.

He chose mathematics for its openness to theorizing, and then made it a practice to read widely over many disciplines. "I went where the ideas led me," he says, explaining that "an idea doesn't care who has it." Woodbury's openness has had results. He's the pioneer of Grade of Measurement, GoM for short, a statistical tool that allows scientists to analyze several categories of data at one time. Appropriately enough, the model can be applied to problems in a wide array of disciplines, from psychology to economics.

In Woodbury's view, humanity owes its relative success to brainpower. "I really believe it's 'use it or lose it.' So I think all the time." Before this interview, for example, he was attacking the problem of why his GoM method behaves very well for up to seven "pure types" but presents problems when more types are analyzed.

Thus began a meditation on the number 7, and how most people can remember seven digits of a telephone number before things get tricky. A rare individual can handle fourteen data items, but that person is probably using both halves of his brain at once. "So how can I invent a methodology to get more data into the analysis? There's my thought for you." Woodbury smiles, perhaps a little bit closer to a solution.

Seven is important to him for one other reason. At that age, wandering about his neighborhood, he came across a group of teenagers who were talking about the deaths of people they knew. Young Woodbury was horrified, but he kept listening. Then one of the boys turned to him and said, "You're going to die, too." Woodbury went home and made a promise to himself: He might die eventually, but he would at least live to see the new millennium.

By Woodbury's calculation, the new millennium begins on January 1, 2001. Should he live to see it, he'll immediately substitute another goal: Hewants to live to see the moment when the Earth, Venus, and the Sun align on June 8, 2004, for the first time since 1882. It's a milestone any true scientist of ideas should observe.

-Eric Larson

Share your comments

Have an account?

Sign in to comment

No Account?

Email the editor