Why do we age?

Article metrics


The evolutionary theory of ageing explains why ageing occurs, giving valuable insight into the mechanisms underlying the complex cellular and molecular changes that contribute to senescence. Such understanding also helps to clarify how the genome shapes the ageing process, thereby aiding the study of the genetic factors that influence longevity and age-associated diseases.


Ageing is usually defined as the progressive loss of function accompanied by decreasing fertility and increasing mortality with advancing age. Such a trait, which impairs survival and fertility, is clearly bad for the individual, raising intriguing questions about why and how it has evolved1,2,3,4. Ageing shows a broad phylogenetic distribution but is not universal, as some species show no age-associated increase in mortality or decline in fertility5. Thus, ageing cannot be explained simply as the inevitable result of biological wear-and-tear. So, why does it occur?

An early explanation for evolution of ageing was the idea that senescence is programmed in order to limit population size or accelerate the turnover of generations, thereby aiding the adaptation of organisms to changing environments. One essential flaw in this argument is that for most species, other than those like Pacific salmon where death coincides directly with the end of a semelparous (once-only) reproductive cycle, there is scant evidence that senescence contributes significantly to mortality in the wild. Natural mortality, as opposed to that seen in protected populations, is mostly due to extrinsic hazards, such as infection, predation, starvation or cold, and occurs mainly in young individuals. As a rule, wild animals simply do not live long enough to grow old (Fig. 1a). Therefore, natural selection has limited opportunity to exert a direct influence over the process of senescence. Even in species where senescence does make some contribution to mortality in the wild (for example, larger mammals and some long-lived birds), any hypothetical 'accelerated ageing gene' would be disadvantageous to the individual. It is therefore difficult to see how genes for accelerated ageing could be maintained in stable equilibrium, as individuals in whom the genes were inactivated by mutation would enjoy a selection advantage.

Figure 1: Evolutionary theories of ageing.

a, Extrinsic mortality in wild environments occurs to an extent that senescence-associated mortality is rare, undermining any idea that genes specifically for ageing have evolved. b, The 'selection shadow' at older ages may permit an accumulation of late-acting deleterious mutations (mutation-accumulation theory). c, Pleiotropic genes that benefit organisms early in life will be favoured by selection even if they have bad effects at later ages (pleiotropy theory). d, Selection pressure to invest metabolic resources in somatic maintenance and repair is limited; all that is required is to keep the organism in sound condition for as long as it might survive in the wild (disposable-soma theory).

The rarity of aged animals in the wild in fact gives the clue to an important principle underlying all of the current evolutionary theories of ageing. As a result of extrinsic mortality, there is a progressive weakening in the force of selection with increasing age6. By an age when wild survivorship has declined to very low levels, the force of selection is too weak to oppose the accumulation of germ-line mutations with late-acting deleterious effects7 (Fig. 1b). This 'selection shadow' allows a wide range of alleles with late deleterious effects to accumulate over the generations with little or no check. This is the 'mutation accumulation' theory, and because the deleterious alleles are essentially unselected, we might expect considerable heterogeneity in the distribution of such alleles among individuals within the population8.

A second theory is that of 'pleiotropy', also sometimes called 'antagonistic pleiotropy'. Williams9 suggested that pleiotropic genes with good early effects would be favoured by selection even if these genes had bad effects at later ages (Fig. 1c). Because the contribution to fitness is a composite of both the size of the effect and the probability of surviving to be affected by it, a small beneficial effect early in life can outweigh a late deleterious effect even if the latter results in senescence and death. This introduces the important idea of a life-history trade-off, which is also a central feature in the third theory, the 'disposable soma' theory10,11, which is based on optimal allocation of metabolic resources between somatic maintenance and reproduction.

Effective somatic maintenance is required only to keep the organism in sound physiological condition for as long as it has a reasonable chance of survival in the wild (Fig. 1d). For example, because more than 90% of wild mice die in their first year12, any investment in mechanisms for survival beyond this age benefits at most 10% of the population. Nearly all of the mechanisms required to combat intrinsic deterioration (such as DNA repair or antioxidant systems) require metabolic resources. Resources are scarce, as shown by the fact that the main cause of mortality for wild mice is cold, owing to failure to maintain thermogenesis13. The disposable-soma theory therefore suggests that the mouse will benefit by investing any spare resource into thermogenesis or reproduction, rather than into better repair capacity, even though this means that damage will eventually accumulate to cause ageing. Although the distinction between the pleiotropy and disposable-soma concepts is sometimes blurred, the latter can be viewed as focusing specifically on mechanisms, particularly the role of somatic maintenance and repair, whereas the former is formulated in terms of a general pattern of gene action and may involve pleiotropic genes of various kinds.

The three theories provide complementary explanations for why ageing occurs. Each also addresses the question: why do species have the life spans they do? The principal determinant in the evolution of longevity is predicted to be the level of extrinsic mortality. If this level is high, life expectancy in the wild is short, the force of selection attenuates fast, deleterious gene effects accumulate at earlier ages, and there is little selection for a high level of somatic maintenance. Consequently, the organism is predicted to be short lived even when studied in a protected environment. Conversely, if the level of extrinsic mortality is low, selection is predicted to postpone deleterious gene effects and to direct greater investment in building and maintaining a durable soma.

Evolutionary theories tested

The evolutionary ageing theories implicitly assume an age-structured population6,9, that is, a population in which individuals can be segregated by age, and therefore predict that ageing should not arise in populations where age classes cannot be assigned. At a unicellular level, it is therefore unsurprising that ageing is generally not seen in bacterial populations. Some unicellular organisms show asymmetry of cell division, such as the budding yeast, Schizosaccharomyces cerevisiae. Mother yeast cells age in that they show an increasing probability of cell death with successive divisions. In multicellular organisms, the need for age structure also applies and ageing is generally predicted to require a clear separation between germ line and soma1,9. The presence or absence of ageing is sometimes attributed to the presence or absence of sexual reproduction, but this is erroneous. It is the distinction between soma and germ line (a common but not universal correlate of sex) that holds the key. In accord with these predictions, two oligochaete species that reproduce by symmetrical fission were found to show no increase in age-specific mortality, whereas four species (two rotifers, one ostracod and one cladoceran crustacean) which reproduce by asexual egg production all showed highly significant increases14. Hydra , which can reproduce sexually but usually reproduces by asexual budding, and which can regenerate a new individual from almost any part of the organism, lacks clear separation of germ line and soma and shows no obvious signs of intrinsic senescence15.

An important prediction of the evolutionary theories is that altering the rate of decline in the force of natural selection will lead to the evolution of a concomitantly altered rate of ageing. This has been tested by applying artificial selection on life-history variables or by intra- and interspecies comparisons of populations that are subject to different levels of extrinsic mortality.

Most selection experiments have used the fruitfly Drosophila melanogaster . By restricting reproduction to later ages, the intensity of selection on the later portions of the life span was increased. This consistently extended the longevity of the selected populations16,17,18,19. Furthermore, a general correlate of delayed senescence has been reduced fecundity in the long-lived flies (Table 1), which supports the idea of a trade-off between fertility and survival, as suggested by the disposable-soma and pleiotropy theories. A similar trade-off was observed in an experiment selecting directly for longevity20. Lastly, when selection operates through the level of extrinsic mortality, populations subjected to low mortality showed increased longevity, longer development times and decreased early fecundity21.

Table 1 Life-history traits and selection for longevity in D. melanogaster

The general finding from selection experiments in Drosophila is therefore that retarded ageing is associated with depression of fitness components in early life, although there is variation in which fitness components are affected. Some studies found that body size and development time were increased in the longer-lived lines21 whereas other studies fail to observe these effects. Sgró and Partridge22 demonstrated that reduced early fecundity — the most consistently observed feature — was causally involved in the retarded ageing of their long-lived lines; after abolishing reproduction through either irradiation or genetic manipulation they found that the differences in ageing rate between their controls and long-lived selected lines disappeared.

Trade-offs have also been reported in the nematode Caenorhabditis elegans where a host of long-lived mutants has been identified. However, although some studies report the same sort of trade-offs between longevity and early-life fitness components as found in Drosophila23, this is not invariably true even for the same genotypes in other laboratories24. A recent study of the long-lived age-1 mutant in C. elegans showed that the relative fitness effects of mutations can be strongly affected by environment25. When mutants were reared together with wild-type individuals under standard culture conditions, neither genotype exhibited a competitive advantage. However, when cultures were alternatively fed and starved — mimicking conditions in nature — the wild type quickly outcompeted the mutant.

Although abundant data support the existence of life-history trade-offs, evidence for the mutation-accumulation theory remains more controversial26. An early Drosophila selection experiment indicated a role for mutation accumulation acting alongside trade-offs27, although probably in a minor capacity. Other studies have directly tested the prediction that mutation accumulation should be revealed by an increase in additive genetic variance in mortality rate at later ages. Initial studies in Drosophila 28,29 seemed to support mutation accumulation but later experiments and further analysis have led to doubts about this conclusion30,31.

From the comparative perspective, numerous opportunities exist to test the prediction that in safe environments (those with low extrinsic mortality) ageing will evolve to be retarded, whereas ageing should evolve to be more rapid in hazardous environments. Adaptations that reduce extrinsic mortality (for example, wings, protective shells or large brain) are generally linked with increased longevity (in bats, birds, turtles and humans). Field observations comparing a mainland population of opossums subject to significant predation by mammals, with an island population not subject to mammalian predation, found the predicted slower ageing in the island population32. Among social insect species, those with the most protected nests contain reproductive females with by far the longest life spans33. A comparative analysis of mortality patterns among birds found that the rate of mortality increase with age was directly correlated with the magnitude of presenescent mortality34.

At the molecular and cellular levels, the disposable-soma theory predicts that the proportional effort devoted to cellular maintenance and repair processes will vary directly with longevity. Numerous studies support this idea. For instance, the long-lived rodent species Peromyscus leucopus exhibits lower generation of the reactive oxygen species (ROS), which are widely seen as an important contributor to the ageing process, higher cellular concentrations of some antioxidant enzymes, and overall lower levels of protein oxidative damage than the shorter-lived species Mus musculus35. A direct relation between species longevity and rate of mitochondrial ROS production in captive mammals has also been found36,37, as has a similar relationship between mammals and similar-sized but much longer-lived birds38. Markers of glycoxidation, the non-enzymatic modification of reducing sugars, are also found to accumulate more slowly in long-lived, as opposed to short-lived, mammals39. DNA repair capacity has been shown to correlate with mammalian life span in numerous comparative studies40, as has the level of poly(ADP-ribose) polymerase41, an enzyme that is important in the maintenance of genomic integrity. The quality of maintenance and repair mechanisms may be revealed by the capacity to cope with external stress. It is notable, therefore, that environmental or genetic manipulations that confer increased longevity on a range of animals also confer increased resistance to environmentally imposed stressors (Table 2). Likewise, comparisons of the functional capacity of cultured cells to withstand a variety of imposed stressors have shown that cells taken from long-lived species have superior stress resistance to that of cells from shorter lived-species8,42,43. All of these studies support the idea that it is the evolved capacity of somatic cells to carry out effective maintenance and repair that governs the time taken for damage to accumulate to levels where it interferes with the organism's viability, and hence regulates longevity.

Table 2 Increased stress resistance in long-lived populations

Specialized life histories

Many organisms live their lives in highly variable environments. In such circumstances we can expect the 'genetic architecture' of the life history, that is, the co-adapted set of traits influencing survival and fecundity, to possess a degree of evolved plasticity that permits a range of optimal responses suited to different circumstances. In poikilotherms, for instance, environmental temperature often influences longevity. This is generally a straightforward consequence of altering metabolic rate, as indicated by the fact that longevity can be similarly altered by manipulating activity44, although the capacity to make such adjustments and retain viability will reflect the range of temperatures to which organisms have been exposed in their evolutionary past. More revealing are cases of plasticity induced by conditions such as lack of food, to which many organisms seem to have evolved a non-reproductive, highly stress-resistant state. Periods of famine often trigger what appear to be metabolic switches that paradoxically extend the normal life span, without sacrificing subsequent reproduction and survival when favourable conditions return45.

The most thoroughly investigated instance of this type of life-history plasticity is seen in the nematode C. elegans. At 20 °C, wild-type C. elegans hermaphrodites live for an average of about 17 days with a maximum of about 25 days. However, under conditions of high larval density and low food availability, larvae develop into the alternative, non-feeding, non-reproducing third instar called the dauer, which can survive for at least 60 days46. If conditions improve, dauers moult to adulthood and exhibit normal adult longevity and reproduction. Although dauers exhibit reduced activity and metabolic rate compared with non-dauer larvae46,47, the preservation of full adult longevity after even an extended dauer period indicates that the effect is not entirely due to reduced metabolism. Like longer-lived strains and species of other animals, dauers are also resistant to a variety of environmental stresses including ROS, temperature extremes and ionizing radiation48. Increased longevity of adult worms results from mutations in several of the genes involved in the dauer pathway, indicating that some activation of dauer physiology may be involved in these mutants, even in the absence of a triggering shortage of food. Exactly what fraction of the longevity-enhancing effect of dauer formation or long-lived genetic mutants is due to reduced metabolic rate is the subject of current controversy23,49.

It is well known that reduced calorie intake slows ageing in laboratory rodents50, an effect that may have parallels with the invertebrate phenomenon. That is, the rodent caloric-restriction response may be a dauer-like adaptive physiological state to 'wait out' periods of food shortage and is generally associated with a partial or complete interruption of fertility. Like many invertebrates, laboratory rodents under caloric restriction show enhanced resistance to a range of stresses51, but unlike invertebrates, no general reduction in mass-specific metabolic rate is required for the effect. Despite the lack of reduced metabolic rate under caloric restriction, Walford and Spindler52 have pointed out a number of biochemical similarities between the caloric-restriction state and that of mammalian hibernation, which can also extend life span. A recent evolutionary model53 has shown that rodents may well have evolved a response to temporary fluctuations in resource availability, in which energy is diverted from reproduction to maintenance functions in periods of food shortage, thereby enhancing survival and retaining reproductive potential for when conditions improve.

Perhaps the most pronounced impact of environment on ageing rate is illustrated by differences between longevity of queens and workers among species of the eusocial insects5. Life spans of workers are typically measured in weeks, those of queens in years. Differences in longevity as great as 100-fold have been reported, even though both queens and workers develop from eggs laid by the same mother and fertilized by the same father. The divergence in their relative longevity is mediated by pheromone exposure or the rate and composition of food supplied to the larvae. Although workers are more physically active, queens produce hundreds of thousands of eggs per year, so the difference in ageing rate is unlikely to be due to differences in metabolic expenditure. Indirect evidence indicates that neuroendocrine factors are involved5, although underlying these factors must be differences in gene expression. A question as yet unanswered is the extent to which such marked differences in mortality schedules result from purely extrinsic factors (danger to the workers foraging in the environment compared with the relative safety of the queen ensconced in a protected, climate-controlled nest surrounded by self-sacrificing defenders) or from differences in the rate of internal decay. This is an area ripe for research into how differential expression in the same configuration of genes might profoundly affect senescence.

Reproduction and ageing

Most discussion of the evolution of ageing focuses on its effects on mortality, rather than reproduction, in spite of the fact that in terms of an impact on fitness, the form of the reproductive schedule is as important as that of the mortality curve54,55. A good reason for this is that many aspects of reproductive senescence can be explained in the same general terms as physiological senescence. Nevertheless, in addition to the important issues of the trade-offs between survival and fecundity, considered earlier, there are intriguing evolutionary questions about the links between reproduction and ageing, notably the significance of post-reproductive survival (where it occurs) and the effects of damage and selection on the germ line.

The existence of a distinct post-reproductive phase is characteristic of certain semelparous species, in which individuals reproduce only once. On the other hand, many semelparous species undergo extremely rapid senescence on completion of reproduction, often as a direct consequence of the massive physiological changes associated with an explosive reproductive burst5. The evolutionary basis of semelparity is well understood54, and from the perspective of the evolutionary theories of ageing, semelparity represents an extreme version of the decline in the force of natural selection with age1. In a semelparous species, the force of natural selection approximates a step function, being uniformly high until reproduction begins, and declining abruptly as reproduction is completed, because the chance of surviving to breed again is effectively zero. This explains the sudden collapse of any pressure to invest in somatic maintenance and repair. Whether or not there is significant post-reproductive survival may be governed chiefly by whether or not the post-reproductive adult contributes actively to the survival chances of the offspring.

A very different example of post-reproductive survival is the human menopause, where fertility in iteroparous human females comes to a relatively abrupt halt at around the age of 45–50 years, when the impact of senescence on most other functions is still small. Although the proximate cause of menopause seems to be oocyte depletion (linked also with neuroendocrine changes), this begs the question why natural selection has not produced a store of oocytes which lasts for longer. One possibility is that during most of humanity's evolutionary history, women rarely survived beyond 45–50 years, so selection simply produced about as many oocytes as would be required. But evidence from hunter–gatherer communities indicates that even though average life expectancy is short, women who avoid the hazards of early life and reach childbearing age have a reasonable chance of surviving to the age of menopause and beyond56. This indicates that the menopause may have a deeper evolutionary significance.

Early female reproductive senescence has been reported in other species (for example, chimpanzees, macaques and toothed whales) but is generally less clear-cut, indicating that if the menopause has an evolutionary basis, this may be found in the special circumstances of the human life history. In particular, menopause could be linked with the evolution of human longevity, notably, through the effects of increased brain size and sociality9,56,57,58,59,60,61. Increased neonatal brain size coupled with the constraint on the birth canal imposed by the mechanics of a bipedal gait has made giving birth unusually difficult for human females. The risks of childbearing, particularly in the absence of modern obstetric care, would increase even more steeply with age if fertility were to persist during the later period of the life span. The problem of a large brain size is also reflected in the fact that human infants are born unusually early, relative to other species, with respect to the completion of brain growth and development. Infants remain highly dependent for extended periods and, in the ancestral environment, their survival will have been unlikely if their mother died in childbirth. There may thus be a fitness advantage in limiting reproduction to ages when it is comparatively safe, thereby increasing the likelihood of the mother surviving to raise her existing offspring to independence. In addition, post-menopausal women may contribute to the successful rearing of their grandchildren, by providing assistance to their own adult offspring and thereby increasing their inclusive fitness, that is, their overall genetic contribution to future generations. It is likely that a combination of all of these factors is required to explain the human menopause62, which may account for the lack of evolutionary support for menopause in other species63.

Human reproductive ageing also highlights the interesting puzzle that although the germ line must, in a fundamental sense, be immortal (as damage cannot be permitted to accumulate across generations without immediate risk of extinction), there is clear evidence that individual germ cells do accumulate faults. Indeed, it would be surprising if the germ line was immune to accumulation of damage, because germ cells are subject to the same kinds of molecular damage as somatic cells. From a statistical perspective it is clear that the germ-cell population does undergo significant ageing. In the case of the human ovary the rate of follicular loss accelerates from around age 35 years, and male fertility begins to decline from around age 30. There is also an increase in the frequency of chromosomal abnormalities in newborn children as a function of maternal and, to a lesser extent, paternal age64,65. Nevertheless, healthy children born to older parents are not prematurely aged, although there is some suggestion that daughters' (but not sons') longevity is adversely affected by advanced paternal age65. Thus, either germ cells are endowed with special maintenance and repair systems10 — the enzyme telomerase being a good example — or selection at the cell or embryo level during gametogenesis, conception and pregnancy serves to screen out most faults. It may be relevant that during each human menstrual cycle about 20 ovarian follicles are triggered to start the process of maturation, although usually only one completes its development and is ovulated. A mechanism of probable significance in the evolution of female germ-line immortality is the stringent bottleneck in the size of the cellular mitochondrial population in early embryogenesis. A healthy complement of mitochondria is essential for subsequent viability of the offspring, and mutations in mitochondrial DNA (mtDNA) tend to accumulate with age66. If the mitochondrial population in the oocyte contains a fraction of organelles bearing mtDNA mutations, a bottleneck coupled with an effective quality screen might select embryos that carry only intact mitochondria67.


Understanding the forces that have sculpted our genetic makeup is likely to provide important insights that can not only guide our investigation of the molecular and cellular basis of ageing, but may also help to identify new routes to positive interventions in the ageing process.

There is a clear prediction that multiple genes influence the ageing process and that these are of several kinds (Box 1). The basis of genetic variation for longevity has begun to be studied in Drosophila by estimating quantitative genetic parameters and mapping quantitative trait loci68. Gene-expression profiling of ageing rodent tissues is beginning to reveal genes that alter their expression with advancing age or whose expression is altered by interventions, such as caloric restriction, which affect rate of ageing69. Unsurprisingly, several of these genes are being identified as those involved in damage/stress-response pathways. Just possibly, such techniques might also reveal late-acting deleterious alleles, as suggested by the mutation-accumulation theory. In general, however, the evolutionary theories caution against interpreting genes whose expression alters in old age as genes that accelerate ageing, and such changes in gene expression are likely to be secondary consequences rather than primary causes of the ageing process.

An important corollary of the prediction (and emerging evidence) that key genes regulating rate of ageing are those that control somatic maintenance and repair, is that at the level of the individual there is considerable scope for the action of stochastic chance70. Not only will individual cells within tissues experience different random accumulations of faults, but there may also be important stochastic variations in developmental processes resulting, for example, in different numbers of cells being formed in key organs, such as the hippocampus. Variation in initial cell number and damage rate will in turn affect the time taken before a threshold for dysfunction is crossed during the progressive neurodegeneration that occurs later in life. This means that genetically identical individuals, maintained in uniform environments, may nevertheless exhibit considerable variation in aspects of the senescent phenotype, as has been frequently observed in ageing studies on inbred laboratory organisms70. Heterogeneity in the senescent phenotype, arising from intrinsic stochasticity as well as genetic and environmental variations, may also help to explain the intriguing phenomenon that in several species age-specific mortality rate eventually slows its rate of increase and may even decline71. Heterogeneity explains such an effect if we assume that the frailer individuals die first, leaving a residual population that, as time goes by, represents a dwindling sub-population made up of those individuals that were always the most resilient.

In conjunction with exciting advances in genome analysis, there is clearly much scope for further development and testing of the evolutionary theories of why we age.


  1. 1

    Kirkwood, T. B. L. & Cremer, T. Cytogerontology since 1881: a reappraisal of August Weismann and a review of modern progress . Hum. Genet. 60, 101–121 (1982).

  2. 2

    Rose, M. R. Evolutionary Biology of Ageing (Oxford Univ. Press, New York, 1991).

  3. 3

    Austad, S. N. Why We Age (Wiley, New York, 1997).

  4. 4

    Kirkwood, T. B. L. Time of Our Lives: The Science of Human Ageing (Oxford Univ. Press, New York, 1999).

  5. 5

    Finch, C. E. Longevity, Senescence and the Genome (Chicago Univ. Press, Chicago, 1990).

  6. 6

    Charlesworth, B. Evolution in Age-Structured Populations (Cambridge Univ. Press, Cambridge, 1994).

  7. 7

    Medawar, P. B. An Unsolved Problem of Biology (Lewis, London, 1952 ).

  8. 8

    Martin, G. M., Austad, S. N. & Johnson, T. E. Genetic analysis of aging: role of oxidative damage and environmental stresses. Nature Genet. 13, 25–34 (1996).

  9. 9

    Williams, G. C. Pleiotropy, natural selection and the evolution of senescence. Evolution 11, 398–411 ( 1957).

  10. 10

    Kirkwood, T. B. L. Evolution of ageing. Nature 270, 301– 304 (1977).

  11. 11

    Kirkwood, T. B. L. Human senescence. BioEssays 18, 1009– 1016 (1996).

  12. 12

    Phelan, J. P. & Austad, S. N. Natural selection, dietary restriction and extended longevity. Growth Dev. Aging 53, 4–6 (1989).

  13. 13

    Berry, R. J. & Bronson, F. H. Life history and bioeconomy of the house mouse. Biol. Rev. 67, 519– 550 (1992).

  14. 14

    Bell, G. Evolutionary and nonevolutionary theories of senescence. Am. Nat. 124, 600–603 ( 1984).

  15. 15

    Martinez, D. E. Mortality patterns suggest lack of senescence in hydra. Exp. Gerontol. 33, 217–225 ( 1997).

  16. 16

    Rose, M. R. & Charlesworth, B. A test of evolutionary theories of senescence. Nature 287, 141– 142 (1980).

  17. 17

    Luckinbill, L. S., Arking, R., Clare, M. J., Cirocco, W. C. & Buck, S. A. Selection of delayed senescence in Drosophila melanogaster. Evolution 38, 996– 1003 (1984).

  18. 18

    Partridge, L., Prowse, N. & Pignatelli, P. Another set of responses and correlated responses to selection on age of reproduction in Drosophila melanogaster. Proc. R. Soc. Lond. B 266, 255–261 (1999).

  19. 19

    Buck, S., Vettraino, J., Force, A. G. & Arking, R. Extended longevity in Drosophila is consistently associated with a decrease in larval viability. J. Gerontol. A 55, B292–B301 (2000).

  20. 20

    Zwaan, B. J., Bijlsma, R. & Hoekstra, R. F. Direct selection on lifespan in Drosophila melanogaster . Evolution 49, 649– 659 (1995).

  21. 21

    Stearns, S. C., Ackermann, M., Doebeli, M. & Kaiser, M. Experimental evolution of aging, growth, and reproduction in fruitflies. Proc. Natl Acad. Sci. USA 97, 3309– 3313 (2000).

  22. 22

    Sgró, C. M. & Partridge, L. A delayed wave of death from reproduction in Drosophila. Science 286, 2521–2524 (1999).

  23. 23

    Van Voorheis, W. A. & Ward, S. Genetic and environmental conditions that increase longevity in Caenorhabditis elegans decrease metabolic rate. Proc. Natl Acad. Sci. USA 95, 11399–11403 (1999).

  24. 24

    Johnson, T. E. & Hutchinson, E. W. Absence of strong heterosis for life span and other life history traits in Caenorhabditis elegans. Genetics 134, 465– 474 (1993).

  25. 25

    Walker, D. W., McColl, G., Jenkins, N. L., Harris, J. & Lithgow, G. J. Evolution of lifespan in C. elegans. Nature 405, 296– 297 (2000).

  26. 26

    Stearns, S. C. & Partridge, L. in Handbook of the Biology of Aging (eds Masoro, E. J. & Austad, S. N.) (Academic, San Diego, in the press).

  27. 27

    Service, P. M., Hutchinson, E. W. & Rose, M. R. Multiple genetic mechanisms for the evolution of senescence in Drosophila melanogaster. Evolution 42, 708–716 (1988).

  28. 28

    Hughes, K. A. & Charlesworth, B. A genetic analysis of senescence in Drosophila. Nature 367, 64– 66 (1994).

  29. 29

    Charlesworth, B. & Hughes, K. A. Age-specific inbreeding depression and components of genetic variance in relation to the evolution of senescence. Proc. Natl Acad. Sci. USA 93, 6140–6145 (1996).

  30. 30

    Promislow, D. E. L., Tatar, M., Khazaeli, A. A. & Curtsinger, J. W. Age-specific patterns of genetic variation in Drosophila melanogaster. I. Mortality. Genetics 143, 839– 848 (1995).

  31. 31

    Shaw, F. H., Promislow, D. E. L., Tatar, M., Hughes, K. A. & Geyes, C. J. Toward reconciling inferences concerning genetic variation in Drosophila melanogaster. Genetics 152, 553–566 (1999).

  32. 32

    Austad, S. N. Retarded senescence in an insular population of opossums. J. Zool. 229, 695–708 ( 1993).

  33. 33

    Keller, L. & Genoud, M. Extraordinary lifespans in ants: a test of evolutionary theories of aging. Nature 389 , 958–960 (1997).

  34. 34

    Ricklefs, R. E. Evolutionary theories of aging: confirmation of a fundamental prediction, with implications for the genetic basis and evolution of life span. Am. Nat. 152, 24–44 ( 1998).

  35. 35

    Sohal, R. S., Ku, H.-H. & Agarwal, S. Biochemical correlates of longevity in two closely-related rodent species. Biochem. Biophys. Res. Comm. 196, 7–11 (1993).

  36. 36

    Ku, H.-H., Brunk, U. T. & Sohal, R. S. Relationship between mitochondrial superoxide and hydrogen-peroxide production and longevity of mammalian-species. Free Radical Biol. Med. 15, 621–627 (1993).

  37. 37

    Barja, G. & Herrero, A. Oxidative damage to mitochondrial DNA is inversely related to maximum life span in the heart and brain of mammals . FASEB J. 14, 312–318 (2000).

  38. 38

    Herrero, A. & Barja, G. 8-oxo-deoxyguanosine levels in heart and brain mitochondrial and nuclear DNA of two mammals and three birds in relation to their different rates of aging. Aging Clin. Exp. Res. 11, 294–300 ( 1999).

  39. 39

    Sell, D. R. et al. Longevity and the genetic determination of collagen glycoxidation kinetics in mammalian senescence. Proc. Natl Acad. Sci. USA 93, 485–490 (1996).

  40. 40

    Kirkwood, T. B. L. DNA, mutations and aging. Mut. Res. 219, 1–7 (1989).

  41. 41

    Grube, K. & Bürkle, A. Poly(ADP-ribose) polymerase activity in mononuclear leukocytes of 13 mammalian species correlates with species-specific life span. Proc. Natl Acad. Sci. USA 89, 11759–11763 (1992).

  42. 42

    Ogburn, C. E. et al. Cultured renal epithelial cells from birds and mice: enhanced resistance of avian cells to oxidative stress and DNA damage. J. Gerontol. 53, B287–B292 ( 1998).

  43. 43

    Kapahi, P., Boulton, M. E. & Kirkwood, T. B. L. Positive correlation between mammalian life span and cellular resistance to stress. Free Radical Biol. Med. 26, 495–500 (1999).

  44. 44

    Sohal, R. S. & Donato, H. Effects of experimentally altered life span on the accumulation of fluorescent age pigment in the housefly, Musca domestica. Exp. Gerontol. 13, 335–341 (1978).

  45. 45

    Austad, S. N. Life extension by dietary restriction in the bowl and doily spider, Frontinella pyramitela. Exp. Gerontol. 24, 83– 92 (1989).

  46. 46

    Klass, M. & Hirsh, D. Non-ageing developmental variant of Caenorhabditis elegans. Nature 260, 523–525 (1976).

  47. 47

    Anderson, G. L. Responses of dauer larvae of Caenorhabditis elegans (Nematoda: Rhabditidae) to thermal stress and oxygen deprivation. Can. J. Zool. 56, 1786–1791 (1978).

  48. 48

    Lithgow, G. J. in Handbook of the Biology of Aging (eds Schneider, E. L. & Rowe, J. W.) 55–73 (Academic, San Diego, 1996).

  49. 49

    Vanfleteren, S. L. & De Vreese, A. Rate of aerobic metabolism and superoxide production rate potential in the nematode Caenorhabditis elegans. J. Exp. Zool. 274, 93– 100 (1996).

  50. 50

    Weindruch, R. & Walford, R. L. The Retardation of Aging and Disease by Dietary Restriction (Thomas, Springfield, IL, 1988).

  51. 51

    Masoro, E. J. Influence of caloric intake on aging and on the response to stressors. J. Toxicol. Environ. Health B Crit. Rev. 1, 243– 257 (1998).

  52. 52

    Walford, R. L. & Spindler, S. R. The response to calorie restriction in mammals shows features also common to hibernation: a cross-adaptation hypothesis. J. Gerontol. 52, B179–B183 (1997).

  53. 53

    Shanley, D. P. & Kirkwood, T. B. L. Calorie restriction and aging: a life history analysis. Evolution 54, 740–750 (2000).

  54. 54

    Stearns, S. C. The Evolution of Life Histories (Oxford Univ. Press, New York, 1992).

  55. 55

    Partridge, L. & Barton, N. H. On measuring the rate of ageing . Proc. R. Soc. Lond. B 263, 1365– 1371 (1996).

  56. 56

    Hill, K. & Hurtado, A. M. Ache Life History: the Ecology and Demography of a Foraging People (de Gruyter, New York, 1996).

  57. 57

    Hamilton, W. D. The moulding of senescence by natural selection. J. Theor. Biol. 12, 12–14 ( 1966).

  58. 58

    Kirkwood, T. B. L. & Holliday, R. in The Biology of Human Ageing (eds Collins, A. J. & Bittles, A. H.) 1– 16 (Cambridge Univ. Press, Cambridge, 1986).

  59. 59

    Austad, S. N. Menopause: an evolutionary perspective. Exp. Gerontol. 29, 255–263 (1994).

  60. 60

    Peccei, J. S. The origin and evolution of menopause: the altriciality-lifespan hypothesis . Ethol. Sociobiol. 16, 425– 449 (1995).

  61. 61

    Hawkes, K., O'Connell, J. F., Jones, N. G. B., Alvarez, H. & Charnov, E. L. Grandmothering, menopause, and the evolution of human life histories. Proc. Natl Acad. Sci. USA 95, 1336–1339 ( 1998).

  62. 62

    Shanley, D. P. & Kirkwood, T. B. L. Evolution of the human menopause. BioEssays (in the press).

  63. 63

    Packer, C., Tatar, M. & Collins, A. Reproductive cessation in female mammals. Nature 392, 807–822 ( 1998).

  64. 64

    Risch, N., Reich, E. W., Wishnick, M. M. & McCarthy, J. G. Spontaneous mutation and parental age in humans. Am. J. Hum. Genet. 41, 218–248 ( 1987).

  65. 65

    Gavrilov, L. A. & Gavrilova, N. S. Parental age at conception and offspring longevity. Rev. Clin. Gerontol. 7, 5–12 (1997 ).

  66. 66

    Wallace, D. C. Mitochondrial diseases in man and mouse. Science 283 , 1482–1488 (1999).

  67. 67

    Cummins, J. M. in Female Reproductive Aging (eds Te Velde, E. R., Pearson, P. L. & Broekmans, F. J.) 207–224 (Parthenon, Carnforth, UK, 2000).

  68. 68

    Vieira, C. et al. Genotype-environment interaction for quantitative trait loci affecting life span in Drosophila melanogaster. Genetics 154 , 213–227 (2000).

  69. 69

    Lee, C. K., Klopp, R. G., Weindruch, R. & Prolla, T. A. Gene expression profile of aging and its retardation by caloric restriction . Science 285, 1390–1393 (1999).

  70. 70

    Finch, C. E. & Kirkwood, T. B. L. Chance, Development and Aging (Oxford Univ. Press, New York, 2000).

  71. 71

    Vaupel, J. W. et al. Biodemographic trajectories of longevity. Science 280, 855–860 (1998).

  72. 72

    Service, P. M., Hutchinson, E. W., Mackinley, M. D. & Rose, M. R. Resistance to environmental-stress in Drosophila melanogaster selected for postponed senescence. Physiol. Zool. 58, 380–389 (1985).

  73. 73

    Tatar, M. Evolution of senescence: longevity and the expression of heat shock proteins . Am. Zool. 39, 920–927 (1999).

  74. 74

    Lin, Y. J., Seroude, L. & Benzer, S. Extended life-span and stress resistance in the Drosophila mutant methuselah. Science 282 , 943–946 (1998).

  75. 75

    Migliaccio, E. et al. The p66shc adaptor protein controls oxidative stress response and life span in mammals. Nature 402, 309–313 (1999).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.