In medical school I was taught that the incidence of chronic, disabling disorders, particularly Alzheimer's disease, increases inexorably with age. I therefore expected that people older than 95 years, often called the oldest old, would be my most debilitated patients. Yet when I became a fellow in geriatrics, I was surprised to find that the oldest old were often the most healthy and agile of the senior people under my care. In fact, the morning I was scheduled to interview a 100-year-old man as part of a research project, he told me we would have to delay the visit. He had seen 19 American presidents take office, and he would be busy that morning voting for the next one.

Such encounters made me wonder if the prevailing view of aging as advancing infirmity was partly wrong. Could it be that many people in their upper 90s enjoy good health and that the oldest old constitute a special—and long-misunderstood—population? Since then, the centenarians I have met have, with few exceptions, reported that their 90s were essentially problem-free. As nonagenarians, many were socially engaged and enjoyed the outdoors and the arts. They basically carried on as if age were not an issue. And accumulating evidence indicates that a significant number of the oldest old are indeed healthier than many people in their 80s or early 90s. The common idea that advancing age inevitably leads to extreme deterioration does, indeed, seem to require revision.

Estimated costs of caring for the oldest old in the future might need modification as well. The centenarian population grew by 66 percent in the U.S. from 1980 to 2010. Many demographers predict that 20 million to 40 million people will be aged 85 or older in the year 2040 and that 500,000 to four million will be centenarians in 2050. The economic burden of caring for people older than 85 could be vast, especially if a huge percentage of them need special care. Yet it may well be that health bills for the oldest old will be lower than previously expected.

Some of the first evidence supporting my suspicions came from a study on Alzheimer's disease that I conducted with my mentor, Lewis A. Lipsitz of the Hebrew Rehabilitation Center in Boston. Surveys reported that this disorder devastates the mind and ultimately kills about 40 percent of those aged 85 and older. Some investigators believe that close to 50 percent of 90-year-olds have Alzheimer's disease and that up to 70 percent of centenarians are affected. Many of the studies on which these conclusions are based, however, did not include subjects older than 93 years, which casts some doubt on these projections. In 1991 Lipsitz and I undertook a pilot study to determine whether the occurrence of Alzheimer's at the center, a chronic care hospital, matched the predictions for centenarians. We found that of the 12 residents in their 100s, only four seemed to have Alzheimer's. This low figure—only 33 percent—was particularly striking considering that residents of such facilities are more likely to be impaired than are their counterparts in the community.

Selective Survival
Our finding suggested that, at least cognitively, the oldest old are in better shape than has usually been assumed. What, we wondered, could explain their good condition? We suspect that the answer to this riddle is that, for whatever reason, some people are particularly resistant to acquiring the disorders that disable and kill most people before age 90. Because of this resistance, they not only outlive others, they do so relatively free of disabilities. In a kind of survival-of-the-fittest phenomenon, these individuals seem to be selected for long-term survival because they possess traits that enable them to avoid or delay killer diseases that commonly accompany aging. And if they do incur illnesses, they are better able to deal with them.

The concept of selective survival was applied, somewhat more narrowly, by demographers in the 1970s to older African-American populations. Researchers reported that although the death rates for blacks were higher than for whites in the U.S. up to age 75, the trend reversed after that age. Then, in what some called a crossover phenomenon, whites were more likely to die at a given age than their African-American counterparts were. They speculated that blacks tended to die earlier because more of them were economically disadvantaged and had less access to health care services. Therefore, those who survived represented an unusually vigorous group, able to overcome obstacles that defeated others. Their vigor, in turn, later gave them a survival advantage.

This selective survival hypothesis may also clarify various other once puzzling findings demonstrating unusually good cognitive and physical health in the oldest old. It seems that men who survive into their late 80s become less and less likely to develop Alzheimer's with each passing year. Moreover, the average man in his late 90s has a more intact mind than the average man in his 80s. These patterns probably emerge because men who are susceptible to Alzheimer's generally die of the condition in their 80s or early 90s. These trends would be explained if the group of men who reach their late 90s consist almost exclusively of individuals who are not susceptible to Alzheimer's and who therefore retain their cognitive abilities indefinitely. More study should reveal whether this is the case.

Gender Crossover
Surprisingly, as a group, men older than 90 generally have better mental function than their female peers do. Women with dementia, it seems, tend to live with their illness rather than die from it. As a consequence, very old women on average retain less of their mental abilities than do men of the same age—who represent the healthy survivors left after other men susceptible to dementia have died off.

At later ages, men also do better than women in terms of physical health. Men in their 60s and 70s are more susceptible than are women to strokes and heart attacks. Delayed onset of these acute conditions allows women to survive longer than men. In absolute numbers, many more women are still alive at 95, but in terms of average health, men begin to take the lead. This switch to more mentally and physically fit men after age 90 is called a gender crossover.

Early signs of the gender crossover can be seen in studies of 80-year-olds. Men who survive to this age without major health problems often continue to live without needing special care. Richard M. Suzman and his colleagues at the National Institute on Aging found that men older than 80 years in one such study were more independent than were similarly aged women. Their report indicated that 44 percent of the men in that age group were robust and independent compared with only 28 percent of women. Additionally, Kenneth G. Manton and Eric Stallard of Duke University estimated the active life expectancy—that is, the years of independent life left—for members of the U.S. senior population. Their findings showed that after age 85, men could expect to live a healthy and active life longer than women could.

What biological and environmental factors might allow the oldest old humans to reach age 95 and beyond in good health? Multiple and intertwined influences undoubtedly play important roles. So-called longevity genes seem to protect against the development of diseases; genetically or otherwise determined adaptive abilities enable survivors to avoid potentially life-threatening conditions. Modifications in everyday activities, such as not smoking, practicing better nutrition and exercising, may also help some people stay fit longer. Basic good luck surely helps as well.

The Genetic Factor
Tempting candidates for possible longevity genes in humans are ones that control the body's mechanism for protecting itself against oxygen radicals. These naturally occurring, highly reactive compounds damage DNA and can destroy cells. Everyone has a genetically determined ability to combat this type of damage. Gene variants that give rise to unusually efficient resistance to oxidative damage could well contribute to the life span of the oldest old by slowing the rate at which oxygen radicals damage cells. Other gene candidates include those that repair aging-related damage to DNA or modulate cells' efficiency to produce and use energy.

What is clear, though, is that there is not just one gene or magic bullet responsible for a centenarian's survival advantage. What makes centenarians rare is that they have just the right combinations of likely hundreds of genetic variants that enhance their abilities to age slowly and decrease their risk for age-related diseases. Making matters even more complicated, there are probably many winning combinations associated with different ethnicities and environments. We are encouraged by the fact that a few combinations have an accuracy of more than 80 percent in differentiating centenarians beyond the age of 106 from noncentenarians. This accuracy is consistent with our hypothesis that the influence of genes on survival becomes stronger and stronger with older and older ages beyond 100 years.

Until recently, scientists thought that centenarians not only have genes that protect against biological processes and environmental exposures conducive to aging but also lack genetic variants associated with increased risk of age-related diseases. Three different studies have now shown, however, that centenarians appear to have just as many of the “bad” variants as the general population. Perhaps a key role of the protective variants is to trump these disease-associated variants.

A notable exception is a variant of the gene coding for the protein apolipoprotein E (APOE) that has been tied to a substantially increased risk of acquiring Alzheimer's. There are three common variants: e2, e3 and e4. People who inherit two e4 genes (one from each parent) have eight times as great a risk as the general population of developing the disease, and those who acquire the disease display symptoms at an average age of 68. Alzheimer's patients with two e3 genes demonstrate symptoms of the disease somewhat later, at about 75 years. The role of e2 remains unclear, but there is evidence that it is associated with a lower risk of developing Alzheimer's. As it turns out, the frequency of the e4 variant is very low among centenarians, most likely because it is so strongly associated with Alzheimer's, vascular disease and premature mortality.

Genes may provide the blueprint for how long a person might live. In effect, they can be considered indicators of how well a person can cope with disease. As such, genes help to determine two interrelated properties that influence aging: adaptive capacity and functional reserve. Adaptive capacity is a person's ability to overcome a disease or injury or to cope with such stresses effectively. Functional reserve refers to how much of an organ is required for its adequate performance. Obviously, one's adaptive capacity depends in part on the body's functional reserve because the ability to deal with disease requires the proper functioning of organs.

The importance of these two characteristics to the survival of many oldest old can be seen in the varying effects that the buildup of neurofibrillary tangles has on cognition. Neurofibrillary tangles describe the web of dead brain cells that occur naturally with aging but appear in abundance in patients with Alzheimer's. The number of tangles that can accumulate before signs of Alzheimer's emerge varies. For example, an autopsy revealed that a 103-year-old man who displayed few outward signs of Alzheimer's had a level of neurofibrillary tangles that in a younger brain would indicate the patient was probably demented. Presumably, the older man had an excess reserve of brain function that allowed him to compensate for the process that was damaging his brain. Perhaps people who have a slow buildup of tangles and a high tolerance for them can remain mentally intact for a long time, showing overt signs of Alzheimer's only very late in life, if at all.

New Thoughts on Aging
The discovery that many people older than 95 are in good shape may mean that future planning for the health care of the oldest old will need to be revised. Much of that planning is based on the theory that although lethal conditions might be postponed as medical technology improves, the incidence of degenerative diseases will be unaltered. The theory predicts that the oldest old will keep suffering from more disease and chronic disability than people in their 80s do. If this theory were true, then the continued increase in the size of the oldest old population would portend a significant burden of poor health and quality of life among the oldest members of our society.

The emerging data, however, fit better with an opposing theory. James F. Fries of Stanford University has proposed that healthier ways of life and medical advances will compress morbidity, mortality and disability into a shorter time period. Thus, the onset both of major fatal diseases (heart disease, cancer, stroke and Alzheimer's) and of age-associated debilitating diseases (degenerative joint disease, sensory impairments and benign memory loss) would be postponed.

Consistent with Fries's hypothesis, robust centenarians often have a relatively short period of infirmity before death. Supercentenarians, a term for people who live to 110 years and beyond, are extremely rare. Yet the New England Centenarian Study has enrolled enough of these “super agers” to demonstrate that, on average, they spend only the last five years of their exceptionally long lives with demonstrable disease or disability.

Jeanne Calment of Arles, France, died at age 122 in August 1997, making her the longest living person ever. Most of us with Methuselean aspirations, in contrast, are up against incredible odds. But recent research on the oldest old has prompted new thinking about the biology of aging. Genetic, biochemical and epidemiological studies should reveal why some people possess resistance to debilitating conditions—and may offer ways to increase that ability in a broader swath of the population. Further, to our relief, the research implies that as the oldest old become more numerous, they may not become a massive drain on the economy. Counter to prevalent theories of aging, many people in their late 90s or 100s lead active, healthy lives. If they represent a survival-of-the-fittest cohort, the time may have come to abandon our past perceptions of our oldest citizens.