Secrets of Healthy Aging and Longevity From Exceptional Survivors Around the Globe: Lessons From Octogenarians to Supercentenarians

WE have all heard stories of cigar-chomping hundred-year-olds who drink copious amounts of vodka (or other spirits) and can climb hills or swim laps faster than most fifty-year-olds. Such hardy, long-lived individuals have been a fascination of society since recorded history (1). Unfortunately, finding individuals who fit this robust description—and who possess valid birth certificates—has proven elusive. However, in the past few years, the credible study of exceptional human longevity has blossomed. In particular, more comprehensive study of long-lived individuals who are free of major clinical diseases and disability, and who might be called “exceptional survivors,” is beginning (2). Some researchers believe that by studying “healthy aging” rather than focusing on specific diseases, we might find protective genetic or environmental secrets that will benefit both length and quality of life (3–5). Discovering factors that enhance odds of healthy aging and translating these findings into evidence-based interventions is becoming a research priority.

Early Studies of Exceptional Human Health and Longevity

The study of exceptional human longevity is not new. In fact, it has a colorful history. Among the first organized research on long-lived humans was a study of alleged centenarians, some said to be older than 160 years, which occurred in the early 1970s in the Caucasus region of the old Soviet Union, the Hunza Valley in Pakistan, and the village of Vilcabamba in Ecuador. Leaf (6) reported unusually high prevalence rates of centenarians, many of whom were remarkably healthy and disability free, in a cover story in National Geographic. When these individuals were later found to be much younger than their stated ages—most were nonagenarians at best—a dubious shadow was cast over research on exceptional human longevity (1,7,8). While these studies did reveal a few healthy elderly persons (and certainly helped boost yogurt sales), rampant age exaggeration made any study of healthy aging in these populations of questionable value. Moreover, the primitive state of quantification of aging-related traits, including assessment of morbidity, physical or cognitive function, and disability made such investigations challenging—even if valid ages had been obtainable.

Healthy Aging and the Exceptional Survivor: the New Paradigm

Much has changed. In the past four decades we have seen a blossoming of interest in the topic of healthy aging and longevity. New paradigms emerged including the concepts of “compression of morbidity” (9), active life expectancy (10), “effective aging” (11), and “successful aging” (12). Initially these concepts were largely descriptive, but gerontologists have begun to quantify healthy aging using methods that account for age-dependent and time-dependent change in morbidity (diseases), physical and cognitive function, and disability (4,13–15). At the same time, we have seen genotyping methods evolve from small candidate gene studies to large, complex whole-genome analyses, gene expression, and beyond. This has led to the discovery of so-called “longevity genes” in model organisms (16) and humans (17–20). Undreamed of advances in proteomes, metabolomes, and a plethora of other “omes” are occurring (21).

After an inauspicious start in the early 1970s, centenarian studies now number over a dozen, and more are in the works (22). Indeed, shortly after the initial National Geographic-funded study (6), the first population-based, comprehensive (and ongoing) study of centenarians began in another Shangri-La like locale, the palm-tree-laden island of Okinawa (23). Moreover, for the true believers in geographic longevity “hot spots,” there is good news—a few real contenders do seem to be popping up. For example, higher than expected prevalence of exceptional longevity has been found in unexpected locales such as the aforementioned Japanese island of Okinawa (24), in another fascinating island population (mainly long-lived men) in Sardinia (25), and in Seventh-Day Adventist-rich Loma Linda, California (26,27).

Modern centenarian studies are based on much more stringent criteria for age-validation, include more extensive measurement of aging-related traits, and usually assess an extensive set of risk factors, often collecting blood samples for assessment of other phenotypes and genotypes (1). As larger numbers of centenarians have passed the 110-year mark, a demographic feat more challenging than surviving the first 100 years, new terms have entered the gerontological lexicon. The term “supercentenarian,” which in the past may have referred to our cigar-chomping, Vodka-swilling, mountain-climbing centenarian, now refers to any centenarian who has survived 110-plus years (28). More comprehensive studies of these truly exceptional survivors are also beginning to emerge (29,30), including a novel study in this issue of the Journal (31).

The Delay or Avoidance of Major Clinical Disease and Disability: A Life-Span Perspective

Of great interest to gerontologists is that a subset of the exceptionally aged does seem to delay or avoid major clinical diseases and disability into their 90s or 100s (32–34). The limited autopsy data that exist on such exceptional survivors have added credibility to these remarkable clinical observations. For example, in an autopsy study of a “typical” centenarian woman in Okinawa, Bernstein and colleagues (35) found an absence of coronary artery disease, cancer, stroke, and little evidence for major damage in several organ systems. However, clear signs of “wear and tear” and past exposures had occurred including compression fractures of the spine, silicosis and scarring of the lungs, and amyloidosis in several organ systems. Larger case series of autopsies in other centenarian populations exist with a pathology burden in some individuals that is less than one might expect for 100 years of life (36,37).

In an ideal world, one would like to study aging over the life span in such individuals to more precisely assess risk and protective factors, and rates of change in health, organ function, and physical/cognitive abilities. This, however, would be a very, very long study. Therefore, most studies of exceptional survivors have traditionally had retrospective or “case-control” designs. It has been exceedingly difficult to recruit large numbers of long-lived “cases” and find age-matched, deceased “controls.”

We are, on the other hand, at a unique point in gerontology. Many cohorts that were originally recruited for the study of limited diseases, such as cardiovascular diseases or cancer, have now been followed for several decades. Some very old cohorts, originally recruited for other purposes, now have National Institute on Aging (NIA)-funded studies of aging and longevity, such as the Framingham Study and Honolulu Heart Program (HHP), both funded by the National Heart, Lung, and Blood Institute for many years. The Framingham Study has several aging-related studies including whole-genome analysis of age-related traits (38). The HHP has several studies of aging including the Honolulu Asia Aging Study (HAAS), a study of cognitive aging (39), and the Hawaii Lifespan Study, a study of healthy aging (4). A few rare cohorts that were originally created to study aging, such as the Baltimore Longitudinal Study of Aging (40), are still in existence. These studies are among a select few that have decades of prospectively collected data and contain more than a few nonagenarians and centenarians. For example, the HHP/HAAS study has close to 2,000 nonagenarians and centenarians (current survivors range in age from 89–107 years) who enrolled at middle age and have been followed for over 40 years (4).

Thus, while the longitudinal study of Exceptional Survival (ES) has been limited (2,41,42), partly due to lack of appropriate datasets, the aforementioned studies are among a relatively small group of studies that performed quantitative phenotyping of age-related traits at multiple time points over the life span. The collection of extensive risk factor data at baseline (and several other time points) enables assessment of risk and protective factors for healthy aging from a life-span perspective. Understanding how the exceptional survivors in such cohorts maintain health until very late in life could help identify life-course pathways that enable more of us to live longer, healthier, and fuller lives (2,41).

Defining the Healthy Aging Phenotype: What is an Exceptional Survivor?

Yet, despite advances in gerontological research methods, one of the greatest challenges in understanding healthy aging is the lack of precise phenotypes (41). What is healthy aging and how do we measure it? There are no universally agreed-upon markers of biological aging (43) and no set “healthy aging” criteria to use as intermediate phenotypes or outcomes. Measurement of ADLs (activities of daily living) is among the most commonly used gerontological assessment tools. Yet, even this is not fully standardized. Some studies ask whether study participants receive help, others assess how “difficult” the task is, some ADL assessments are done over the telephone, others by direct observation. The scale itself is usually an ordinal ranking of ability (10). Different studies, different countries, and different cultures approach the study of healthy aging in different, often nonquantitative, ways. So much so that the utility of the concept of healthy or “successful” aging has been questioned (44).

This is beginning to change. In 2001, five NIA programs organized the Advisory Panel on Exceptional Longevity (APEL) (41) in order to assess the emerging opportunities for quantitative study of healthy aging and longevity in humans. APEL made several recommendations for the study of genetic and other factors that contribute to exceptional longevity. APEL also reported that experience is limited in studies of exceptional human longevity, and stated that the single most important prerequisite in planning such studies is phenotype definition.

The APEL report suggested that exceptional longevity can be defined in numerous ways, and the definition or “phenotype” selected will strongly influence the analytic approach. Phenotype based on age alone is obvious and simple. It is not necessarily the ideal. For example, a high age threshold for study participants (such as 100 years), is associated with a high probability of longevity recurrence in siblings (45–47), suggesting strong genetic influence. On this basis many researchers have simply chosen the age of 100 or “centenarianism” as their phenotype, without careful consideration of cohort effects, ethnicity, gender, population admixture, or other confounding factors (41). However, the age threshold chosen should consider several of these criteria, as well as the availability of sufficient study participants for adequate statistical power.

Conversely, an alternative to a predetermined age, such as 100, is to treat survival as a quantitative variable and to study risk or protective factors associated with attained age in a set cohort, such as a birth cohort (41). Appropriate gender, ethnicity, and cohort-specific age percentile cutoffs should facilitate the creation of definitions for ES that control for important confounding factors (41). While various study approaches exist, there was agreement in the APEL report that more longitudinal studies are needed and that there should be more collaboration across disciplines, cohorts, and countries. Some of these collaborations are beginning to occur such as U.S.-based Genetics of Longevity Consortium (48) and the Long Life Family Study (49), and the Europe-based Genetics of Healthy Aging Consortium (50).

Global Approaches to Understanding Healthy Aging: The Role of the Exceptional Survivor

A remarkable increase in the population of very old persons, particularly centenarians, is occurring worldwide mainly due to decreasing mortality rates at older ages. In fact, centenarians are now among the fastest-growing age groups (51). The U.S. Census bureau predicts 834,000 centenarians will exist in the United States by the year 2050, although the prediction has a remarkable range—from 265,000 to 4,200,000 (51). In any case, the numbers will be impressive. The implications of this longevity boom will be daunting. Despite some good news, such as evidence that lifetime medical expenses for acute care (i.e., Medicare, drug costs) actually plateau at very old ages, other costs, such as long-term care, increase substantially since most of the very old need some form of assistance with activities of daily living (ADLs) (52). This should be cause for concern since the oldest-old account for a disproportionate amount of health care costs, and long-term care is a major expense for elderly people (52). Therefore, it is imperative that we better understand frailty, morbidity, and disability in long-lived persons and identify the factors that allow some of these persons to maintain health and independence until the end of their very long lives.

It was in this spirit—a desire to further explore healthy aging and longevity in studies that focus on the exceptional survivor—that the Journal has brought together a diverse group of studies from around the globe. It was also our good fortune that most studies in this special section have prospectively collected data, and that each study has a significant number of exceptional survivors, some of whom have been followed for several decades. The studies presented in this special section address important aspects of ES in different but complementary ways—and even include a few of those elusive supercentenarians.

Three of the studies collected herein address physical and cognitive function in very old age. Engberg and colleagues (53) address disability in two successive cohorts of centenarians in Denmark, finding that women (but not men) had improved ADLs in the 1905 versus 1895 birth cohort. This is consistent with findings of the U.S. National Long Term Care Survey (54) for women but contrary to what has been found in the past for successive rapidly expanding centenarian cohorts in Okinawa, Japan (55). Also in the Journal, but further East, on the main island of Japan, Dodge and colleagues (56) examined the nature and frequency of leisure activities of Japanese elders aged from 65 to near 100 years. They found that all cognitive domains tested were associated with frequency of engagement in “nonphysical” hobbies and that gait speed was a strong predictor of reduction in leisure activities. A third study in the Journal, that of supercentenarians in Okinawa, assessed physical, cognitive, and other phenotypic characteristics of this elite group as they progressed from age 100 to 110-plus years (31). Interestingly, an even more functionally elite phenotype was observed in supercentenarians than typical centenarians (34). Supercentenarians maintained ADL independence, on average, until age 105 years. This is about 5 to 10 years longer than typical centenarians (34), and supports an extended “health span” in such individuals rather than simply gaining extra years of disability. Understanding how well people function at exceptional ages, both physically and cognitively, and the contributing factors would have important public health implications.

This underscores the need for more study of predictors of healthy aging and longevity in the very old. In this vein, two reports, one of Japanese centenarians (57) and one of Swedish centenarians (58) focus on the question of survival predictors in this population. In a fascinating study, Arai and colleagues (57) found that a set of novel risk factors linked to adipose–endocrine function and the insulin-like growth factor-1 axis could reliably predict survival in a very old Japanese population. An innovative mortality risk prediction tool was developed that should be assessed in other populations. Half a world away, in Sweden, in a centenarian cohort followed to extinction, Hagberg and Samuelsson found that baseline physiological reserve and present health and function were the strongest survival predictors in these very old Swedes (58). Interestingly, even at age 100 years, high-density lipoprotein (HDL) was still a predictor of further survival, and red wine drinkers lived longer.

Finally, a growing literature suggests that genetic factors may be particularly important for healthy aging and exceptional longevity. Further addressing this field of inquiry, the final two articles of the Journal also use prospective study designs to assess the potential role of variation in two particular genes and risk for healthy aging and longevity. The first article, that of Delmonico and colleagues (59), examined the association of variation within the alpha-actinin-3 (ACTN3) gene on muscle function and physical performance in older adults. The ACTN3 R577X polymorphism predicted decline in some performance variables with age in men and women. Although the results could not be replicated in another cohort, the search for such gene variants will undoubtedly lead to new and exciting biological targets. The second and final article, that of Koropatnick and colleagues (60), also found that plasma HDL cholesterol, and a particular CETP (cholesteryl ester transfer protein) gene variant that affects HDL levels, and is common in the Japanese population, were predictive of healthy aging and longevity in older Japanese-American men in Hawaii.

Conclusion

The articles presented in this special section of the Journal represent a new and exciting direction in gerontology—a focus on health rather than disease. While some progress has been made in understanding healthy aging, much remains to be done. Phenotypes that reflect healthy aging, which are about as “complex” as they come, need further refinement in order to better understand their genetic basis. This must be accomplished before we can begin to understand the complex array of factors affecting healthy aging and longevity. Such work would be enhanced by standardization of performance-based measures of aging-related traits across populations (14), the development of consortia and/or large multiethnic cohorts for reproduction, and validation of findings and novel approaches for investigating healthy aging across the life span (41,61–63). When such criteria are met, and truly “translational” work on healthy aging is fostered, it may not be long before today’s version of the “exceptional survivor” will be tomorrow’s “typical” senior citizen. If this happens, the cigar-chomping, mountain-climbing supercentenarians of Shangri-La fame may not prove so elusive after all.