Supplement Understanding the Determinants of Exceptional Longevity Thomas Perls, MD, MPH, and Dellara Terry, MD, MPH

Centenarians represent an extreme of life expectancy. They achieve their exceptional longevity in part by lacking genetic variations linked to premature death. Pedigree studies have shown a substantial familial component in the ability to survive to extreme old age, and a recent study demonstrated a locus on chromosome 4 linked to exceptional longevity, indicating the likely existence of at least one longevity-enabling gene in humans. The children of centenarians have markedly reduced relative risks for age-related diseases, particularly heart disease, hypertension, and diabetes,

and are a promising model for genetic and phenotypic studies of 1) aging slowly relative to the general population and 2) the delay of and perhaps escape from important age-related diseases. These studies and those of other mammals and lower organisms show great promise for the delineation of important environmental and genetic determinants of aging well.

R

gene involved, the Werner (WRN) gene, is a DNA helicase that catalyzes DNA unwinding. At the molecular level, persons with the Werner syndrome have substantial genomic instability, with prevalent chromosome breaks, translocations, and large deletions (7). The cause of another progeroid syndrome, the Hutchinson–Gilford syndrome, was recently found to be a single mutation in the Lamin A gene (8, 9). At the other extreme are centenarians, who outlive those who are relatively predisposed to age-related fatal illnesses. Centenarians are less likely to have environmental and genetic exposures that contribute to death at earlier ages. This phenomenon, called demographic selection, is exemplified by the fact that the apolipoprotein E ⑀4 allele, associated with heart disease and Alzheimer disease, is rare in centenarians, whereas the prevalence of an alternative allele, ⑀2, is relatively high (10). Richard Cutler, in what is now a classic paper in gerontology (11), proposed that persons who achieve extreme old age do so in great part because they have genetic variations that affect the basic mechanisms of aging and that result in a uniform decreased susceptibility to age-associated diseases. Our studies and those of others researching the oldest old have noted that persons who achieve extreme old age probably lack many of the variations (the “disease genes”) that substantially increase risk for premature death by predisposing persons to various fatal diseases, both ageassociated and non–age-associated (12). More controversial is the idea that genetic variations might confer protection against the basic mechanisms of aging or age-related illnesses (the “longevity-enabling genes”) (13). The progressive selecting out of more and more genetically fit persons of very old age lays the foundation for a simpler model for sorting out the genetics of aging and longevity. Centenarians may be rare because a complex set of environmental and genetic variables must coexist for such survival to occur.

esearchers and the lay public often ask how much of aging, or of age-related disease, is due to genetic factors, which we generally cannot influence, and how much is due to health-related behaviors and interventions, which we can influence. This question has been addressed from many points of view, ranging from evolutionary biology to genetic epidemiology. Clearly, the more that successful aging is due to environmental factors, the more likely it is that we have the power to determine our health and vitality in old age. On the other hand, if environmental influences explain much of the variability in healthy aging, it seems a daunting task to sort out the many environmental, genetic, and stochastic factors and interactions that affect how we age (1). A Scandinavian study of twins (2) calculated the heritability of average life expectancy to be 20% to 30%. In other words, environmental differences accounted for at least 70% of the variations in age at death for the sets of twins studied, the oldest of whom were in their mid- to late 80s. If average humans are born with an average set of genetic polymorphisms, it will be differences in their habits and environments that explain the variability in their life expectancy. Supporting this idea is a study of Seventh Day Adventists (3) indicating that optimal health-related behaviors add 10 years to average life expectancy. Given that in the United States, 75% of persons are overweight and one third are obese (4), far too many persons still use tobacco (5), and far too few persons regularly exercise (6), it is no wonder that our average life expectancy is about 10 years less than our average set of genes can achieve for us. The heterogeneity of how people age represents a broad spectrum of the importance of environmental, genetic, and stochastic determinants of survival. The persons at the extremes of longevity—persons with a progeroid syndrome and those who live to extreme old age—might achieve those extremes through factors that are, individually or in combination, rare. The progeroid syndromes mimic premature aging in numerous ways, including early development of cataracts, aged skin, and cardiovascular disease. One such condition is the Werner syndrome, the underlying cause of which is a single gene mutation. The

Ann Intern Med. 2003;139:445-449. For author affiliations, see end of text.

THE FAMILIALITY

OF

www.annals.org

EXCEPTIONAL LONGEVITY

To further explore the genetic aspects of exceptional old age, we recently studied 444 centenarian pedigrees containing 2092 siblings (14). We compared sibling death © 2003 American College of Physicians 445

Supplement

Understanding the Determinants of Exceptional Longevity

Table. Relative Survival Probabilities of Siblings of Centenarians Compared with the U.S. 1900 Birth Cohort Age, y

Relative Survival Probability in Men (95% CI)

Relative Survival Probability in Women (95% CI)

20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

1.00 (–) 1.00 (0.99–1.01) 1.01 (1.00–1.02) 1.01 (1.00–1.03) 1.03 (1.01–1.05) 1.04 (1.02–1.06) 1.08 (1.05–1.10) 1.10 (1.07–1.13) 1.18 (1.15–1.21) 1.29 (1.25–1.33) 1.48 (1.42–1.53) 1.68 (1.60–1.77) 2.03 (1.90–2.16) 2.69 (2.47–2.91) 4.08 (3.62–4.54) 8.35 (6.98–9.71) 16.95 (10.84–23.07)

1.00 (–) 1.01 (1.00–1.02) 1.02 (1.01–1.03) 1.03 (1.01–1.04) 1.04 (1.02–1.06) 1.06 (1.04–1.07) 1.08 (1.06–1.09) 1.10 (1.08–1.12) 1.12 (1.09–1.14) 1.16 (1.13–1.19) 1.24 (1.21–1.28) 1.36 (1.31–1.41) 1.54 (1.47–1.60) 1.83 (1.73–1.93) 2.56 (2.39–2.74) 4.15 (3.73–4.57) 8.22 (6.55–9.90)

rates and survival probabilities with national U.S. death rates and survival probabilities according to the Social Security Administration’s life table for the cohort born in 1900. Compared with the 1900 birth cohort, the siblings of centenarians maintained a lifelong reduction in risk for death of approximately one half, even through very old age. Siblings may share environmental and behavioral factors early in life that may have strong effects throughout life. It would make sense that some of these effects are primarily responsible for the shared survival advantage up to middle age. Some of these effects might not become evident until after middle age. However, in general, environmental characteristics—such as socioeconomic status, lifestyle, and region of residence—are likely to diverge as siblings grow older. Thus, if the survival advantage of the siblings of centenarians is mainly due to environmental factors, that advantage should decline with age. In contrast, the stability of relative risk for death across a wide age range suggests that the advantage is due more to genetic than to environmental factors. Whereas death rates reflect the current intensity of death at a moment in time, survival probability reflects the cumulative experience of death up to that moment in a cohort’s life history. Thus, a relatively constant advantage from moment to moment (as seen in the relative death rates) translates into an increasing survival advantage over a lifetime (as seen in the relative survival probabilities). This increase is seen in the Table, which shows the relative survival probabilities of the male and female siblings of centenarians at various ages. By the age of 100 years, the relative survival probability is 8.2 for female siblings of centenarians and 17 for male siblings. From the analysis of death rates, we know that the siblings’ survival advantage does not increase as the siblings age. Rather, the siblings’ relative probability of survival is a cumulative measure and reflects their lifelong 446 2 September 2003 Annals of Internal Medicine Volume 139 • Number 5 (Part 2)

survival advantage over the general population born around the same time. The marked increase in relative survival probability and sustained survival advantage in extreme old age could be consistent with the forces of demographic selection, in which genes or environmental factors (or both) that predispose to longevity win out over those that are associated with premature or average mortality. The substantially higher relative survival probability values for men at older ages might reflect the fact that male mortality rates are substantially higher than female mortality rates at these ages and, thus, that men gain a greater advantage from beneficial genotypes than women do. Another possibility is that men require an even greater, and thus more rare, combination of genetic and environmental factors to achieve extreme age than women do (15). Either possibility could explain why men make up only 15% of centenarians. Although men make up a small proportion of centenarians, they tend to be better off than their female counterparts in terms of both physical and cognitive function (13). At first, this seems paradoxical because so many more women achieve extreme old age. One explanation may be that compared with women, men must be in particularly good condition to achieve extreme old age. Women, on the other hand, seem to be physiologically stronger in old age and more likely than men to be able to live with chronic illnesses and disabilities. These observations may indicate a demographic crossover in which women are better off than men in younger old age and men, although fewer in number, are functionally better off in extreme old age. The underlying reasons that women generally live longer than men and, at least before menopause, are substantially less likely to develop heart disease and stroke are unclear. Estrogen, which has been suggested to be a powerful antioxidant, has been implicated as an important reason (16). Another possible reason is that women, because of menses, are relatively iron deficient compared with men for a 30- to 40-year period (17). Iron is a critical catalyst in mitochondrial production of free radicals as a byproduct of metabolism (18); perhaps a reduction in available iron leads to less free-radical production (19). For example, iron deficiency has been associated with significant reductions in levels of oxidized low-density lipoprotein cholesterol, an essential component of atherosclerotic plaque production (20). In addition, diets high in heme have been associated with significantly increased risk for heart disease (21).

GENES PREDISPOSING

TO

EXCEPTIONAL LONGEVITY

The discovery of genetic variations that explain even 5% to 10% of the variation in survival to extreme old age could yield important clues about the cellular and biochemical mechanisms that affect basic mechanisms of aging and susceptibility to age-associated diseases. Until recently, only one genetic variation had been replicated to show an association with exceptional longevity, but even www.annals.org

Understanding the Determinants of Exceptional Longevity

that finding may vary according to ethnicity and other, unknown sources of stratification. Schachter and colleagues (10) noted that the apolipoprotein E ⑀4 allele becomes markedly less frequent with advancing age. One of its counterparts, the ⑀2 allele, becomes more frequent with advancing age in white persons. Presumably the ⑀4 allele disappears with aging because of its association with “premature” death secondary to Alzheimer disease and heart disease. The fact that just one genetic variation has emerged has made scientists pessimistic that others will be found. However, the elevated relative survival probability values found among the siblings of centenarians nonetheless supported the usefulness of doing genetic studies to determine what genetic region or regions—and ultimately what genetic variations— centenarians and their siblings have in common that confer their survival advantage (22). Centenarian sibships from the New England Centenarian Study were used in a genome-wide sibling-pair study of 308 persons belonging to 137 families with exceptional longevity. According to nonparametric analysis, significant evidence for linkage was noted for a locus on chromosome 4. These results indicate the great likelihood of the existence of a gene or genes that positively influence the ability to achieve exceptional old age (23). The next step will be to replicate this result with an independent set of families and to proceed with a single-nucleotide polymorphism analysis of the locus to find the gene or genes playing an important role in the marked survival advantage of these persons. Pursuing a long, involved, and expensive sibling-pair study prompts the question of how useful it would be to find a gene and polymorphism common to centenarians. Discovering genes that could impart the ability to live to old age while compressing the period of disability toward the end of life would yield important insight into how the aging process increases susceptibility to diseases associated with aging and how this susceptibility might be modulated (24). We anticipate that human longevity-enabling genes will be found to influence aging at its most basic levels, thus affecting a broad spectrum of genetic and cellular pathways in a synchronous manner. The centenarian genome should also be an efficient tool for ferreting out disease genes. Comparison of single-nucleotide polymorphism frequencies of genes implicated in diseases between centenarians and individuals with the diseases should reveal clinically relevant polymorphisms. Another approach that researchers are in the early stages of understanding is differential gene expression in models known to slow the aging process, such as caloric restriction (25). This may prove to be another potent tool for discovering longevityenabling genes. The hope, of course, is that these gene discoveries will help identify drug targets and create drugs that would allow persons to become more “centenarianlike” by maximizing the period of their lives spent in good health. www.annals.org

FUTURE PHENOTYPIC

AND

Supplement

GENETIC DISCOVERIES

The compression of functional impairment toward the end of life seen among centenarians may be consistent with James Fries’s compression-of-morbidity hypothesis (26). Fries proposes that as the limit of human life span is approached, the onset and duration of lethal diseases associated with aging must be compressed toward the end of life. It is this compression that can lead phenotypic studies of centenarians to be studies of frailty rather than studies of phenotypic characteristics the centenarians might have had for most of their lives that were conducive to exceptional longevity. An inherent problem in studying levels of practically anything in centenarians is that most of the persons studied are in the morbidity phase of their lives and have an annual mortality rate of 30% to 50%. Except in truly healthy centenarians, these studies do not reflect levels conducive to living to 100 years of age but rather levels associated with the last year or two of life (27). Although we found that functional impairment was compressed toward the end of life among centenarians, we noted anecdotally that some centenarians had long histories of an age-related disease. Perhaps an unusual adaptive capacity or functional reserve allowed some of these persons to live a long time with what normally would be considered a debilitating, if not fatal, disease while delaying its attendant morbidity and death by as much as decades. To explore this hypothesis in our centenarian sample, we conducted a retrospective cohort study (28) exploring the timing of age-related diseases. Three profiles emerged from the analysis of health history data. Forty-two percent of the participants were “survivors,” in whom an age-associated disease was diagnosed before the age of 80 years. Forty-five percent were “delayers,” in whom an age-associated disease was diagnosed at or after the age of 80 years, beyond the average life expectancy for their birth cohort. Thirteen percent were “escapers,” who attained their 100th birthdays without diagnosis of any of the 10 age-associated diseases studied. The survivor, delayer, and escaper profiles represent different centenarian phenotypes, and probably different genotypes as well. The categorization of centenarians into these and other groupings (for example, cognitively intact persons or smokers without smoking-related illnesses) should prove useful in the study of factors that determine exceptional longevity. Although centenarians may be a scientifically valuable cohort for the discovery of genetic correlates of exceptional longevity, phenotypic measures of centenarians, except perhaps escapers, might be correlates of marked frailty rather than of the ability to achieve extreme old age. However, as two recent studies suggest (29, 30), the children of centenarians seem to be unusually healthy, and it may prove worthwhile to study them for both phenotypic and genetic determinants of the ability to live to 100 years of age and more. Terry and colleagues (29) compared the health histo2 September 2003 Annals of Internal Medicine Volume 139 • Number 5 (Part 2) 447

Supplement

Understanding the Determinants of Exceptional Longevity

ries of 177 unrelated children of centenarians with those of birth cohort–matched controls. The controls had parents who were born in the same years as the centenarians, but at least one of those parents had died at the age of average life expectancy. After multivariate adjusted analyses, the centenarian offspring had reduced relative prevalences of 56% for heart disease, 66% for hypertension, and 59% for diabetes. Thus, the offspring of centenarians showed a markedly reduced prevalence of diseases and conditions associated with aging, particularly cardiovascular disease and cardiovascular risk factors. Barzilai and colleagues (30) studied lipid profiles among Ashkenazi Jewish centenarians, their children, and the children’s spouses (the controls in the study). Both male and female children had significantly higher high-density lipoprotein cholesterol levels compared with controls, and the male children also had significantly lower low-density lipoprotein cholesterol levels than did controls. These two studies support the hypothesis that phenotypic and probably genotypic characteristics conducive to exceptional longevity are transmitted in long-lived families and, furthermore, that factors related to cardiovascular health seem to play a particularly important role.

CONCLUSION The careful phenotyping of numerous animal and human models of aging, the collection of genetic material, and the current explosion in molecular genetics data and techniques are likely to soon fill important gaps in the aging puzzle. Complex gene–gene and gene–environment interactions will certainly complicate our ability to understand how genes affect aging. However, with the power of demographic selection, centenarians have already proven helpful in deciphering some polymorphisms and genetic loci associated or not associated with exceptional old age. The children of centenarians, who seem to be following closely in their parents’ footsteps, might yield additional discoveries about phenotypic and genetic correlates of successful aging. From Boston University Medical Center, Boston, Massachusetts. Acknowledgments: The authors thank the American Association of Re-

tired Persons for assistance with participant recruitment and dissemination of findings, and the centenarians and their family members who participate in the New England Centenarian Study. Grant Support: By the American Federation of Aging Research’s and

Alliance for Aging Research’s Paul Beeson Faculty Scholar in Aging Research Award, the Ellison Medical Foundation, the Institute for the Study of Aging, the Alzheimer’s Association, and the National Institute on Aging (grant RO1 AG18721). Potential Financial Conflicts of Interest: None disclosed. 448 2 September 2003 Annals of Internal Medicine Volume 139 • Number 5 (Part 2)

Requests for Single Reprints: Thomas Perls, MD, MPH, Geriatrics

Section, Boston University Medical Center, 88 East Newton Street, Boston, MA 02118; e-mail, [email protected]. Current author addresses are available at www.annals.org.

References 1. Martin GM. Keynote: mechanisms of senescence— complificationists versus simplificationists. Mech Ageing Dev. 2002;123:65-73; discussion 75-9. [PMID: 11718800] 2. Ljungquist B, Berg S, Lanke J, McClearn GE, Pedersen NL. The effect of genetic factors for longevity: a comparison of identical and fraternal twins in the Swedish Twin Registry. J Gerontol A Biol Sci Med Sci. 1998;53:M441-6. [PMID: 9823748] 3. Fraser GE, Shavlik DJ. Ten years of life: is it a matter of choice? Arch Intern Med. 2001;161:1645-52. [PMID: 11434797] 4. Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. JAMA. 2003;289:187-93. [PMID: 12517229] 5. Wechsler H, Rigotti NA, Gledhill-Hoyt J, Lee H. Increased levels of cigarette use among college students: a cause for national concern. JAMA. 1998;280: 1673-8. [PMID: 9831998] 6. Wei M, Kampert JB, Barlow CE, Nichaman MZ, Gibbons LW, Paffenbarger RS Jr, et al. Relationship between low cardiorespiratory fitness and mortality in normal-weight, overweight, and obese men. JAMA. 1999;282:1547-53. [PMID: 10546694] 7. Yu CE, Oshima J, Fu YH, Wijsman EM, Hisama F, Alisch R, et al. Positional cloning of the Werner’s syndrome gene. Science. 1996;272:258-62. [PMID: 8602509] 8. Eriksson M, Brown WT, Gordon LB, Glynn MW, Singer J, Scott L, et al. Recurrent de novo point mutations in lamin A cause Hutchinson-Gilford progeria syndrome. Nature. 2003;423:293-8. [PMID: 12714972] 9. De Sandre-Giovannoli A, Bernard R, Cau P, Navarro C, Amiel J, Boccaccio I, et al. Lamin A truncation in Hutchinson-Gilford progeria. Science. 17 April 2003. [PMID: 12702809] 10. Schachter F, Faure-Delanef L, Guanot F, Rouger H, Froguel P, LesueurGinot L, et al. Genetic associations with human longevity at the APOE and ACE loci. Nat Genet. 1994;6:29-32. [PMID: 8136829] 11. Cutler RG. Evolution of human longevity and the genetic complexity governing aging rate. Proc Natl Acad Sci U S A. 1975;72:4664-8. [PMID: 812099] 12. Schachter F. Causes, effects, and constraints in the genetics of human longevity. Am J Hum Genet. 1998;62:1008-14. [PMID: 9545419] 13. Perls T, Kunkel LM, Puca AA. The genetics of exceptional human longevity. J Am Geriatr Soc. 2002;50:359-68. [PMID: 12028221] 14. Perls TT, Wilmoth J, Levenson R, Drinkwater M, Cohen M, Bogan H, et al. Life-long sustained mortality advantage of siblings of centenarians. Proc Natl Acad Sci U S A. 2002;99:8442-7. [PMID: 12060785] 15. Perls T, Fretts R. Why women live longer than men. Scientific American Presents. 1998; 100-7. 16. Kannel WB, Hjortland MC, McNamara PM, Gordon T. Menopause and risk of cardiovascular disease: the Framingham study. Ann Intern Med. 1976;85: 447-52. [PMID: 970770] 17. Sullivan JL. The iron paradigm of ischemic heart disease. Am Heart J. 1989; 117:1177-88. [PMID: 2653014] 18. Stohs SJ, Bagchi D. Oxidative mechanisms in the toxicity of metal ions. Free Radic Biol Med. 1995;18:321-36. [PMID: 7744317] 19. Sullivan JL. Iron and the sex difference in heart disease risk. Lancet. 1981;1: 1293-4. [PMID: 6112609] 20. Salonen JT, Korpela H, Nyyssonen K, Porkkala E, Tuomainen TP, Belcher JD, et al. Lowering of body iron stores by blood letting and oxidation resistance of serum lipoproteins: a randomized cross-over trial in male smokers. J Intern Med. 1995;237:161-8. [PMID: 7852918] 21. Klipstein-Grobusch K, Grobbee DE, den Breeijen JH, Boeing H, Hofman A, Witteman JC. Dietary iron and risk of myocardial infarction in the Rotterdam Study. Am J Epidemiol. 1999;149:421-8. [PMID: 10067901] www.annals.org

Understanding the Determinants of Exceptional Longevity 22. McCarthy MI, Kruglyak L, Lander ES. Sib-pair collection strategies for complex diseases. Genet Epidemiol. 1998;15:317-40. [PMID: 9671984] 23. Puca AA, Daly MJ, Brewster SJ, Matise TC, Barrett J, Shea-Drinkwater M, et al. A genome-wide scan for linkage to human exceptional longevity identifies a locus on chromosome 4. Proc Natl Acad Sci U S A. 2001;98:10505-8. [PMID: 11526246] 24. Hitt R, Young-Xu Y, Silver M, Perls T. Centenarians: the older you get, the healthier you have been [Letter]. Lancet. 1999;354:652. [PMID: 10466675] 25. Lee CK, Klopp RG, Weindruch R, Prolla TA. Gene expression profile of aging and its retardation by caloric restriction. Science. 1999;285:1390-3. [PMID: 10464095] 26. Vita AJ, Terry RB, Hubert HB, Fries JF. Aging, health risks, and cumulative

www.annals.org

Supplement

disability. N Engl J Med. 1998;338:1035-41. [PMID: 9535669] 27. Perls T. Genetic and phenotypic markers among centenarians. Journal of Gerontology: Medical Sciences. 2000;56A:M1-4. 28. Evert J, Lawler E, Bogan H, Perls T. Morbidity profiles of centenarians: survivors, delayers and escapers. Journal of Gerontology: Medical Sciences. [In press.] 29. Terry DF, Wilcox M, McCormick MA, Lawler E, Perls TT. Cardiovascular advantages among the offspring of centenarians. Journal of Gerontology: Medical Sciences. [In press.] 30. Barzilai N, Gabriely I, Gabriely M, Iankowitz N, Sorkin JD. Offspring of centenarians have a favorable lipid profile. J Am Geriatr Soc. 2001;49:76-9. [PMID: 11207846]

2 September 2003 Annals of Internal Medicine Volume 139 • Number 5 (Part 2) 449

Current Author Addresses: Drs. Perls and Terry: Geriatrics Section,

Boston University Medical Center, 88 East Newton Street, Boston, MA 02118.

E-450

Annals of Internal Medicine Volume • Number

www.annals.org