Did aging evolve to prevent epidemics?

How did aging evolve? Some evolutionary theories invoke tradeoffs between maintenance/repair and reproduction. Others postulate that genes that cause age-related decline can be positively selected, so long as these same genes confer a fitness advantage early in life.

A common feature of these theories is that they operate at the level of the individual organism, rather than the species. Models based on group selection usually have logical problems. For example, suppose that aging evolved in order to eliminate post-reproductive old organisms to preserve resources for the reproductively competent young. This is circular: Why are the old organisms were post-reproductive in the first place? i.e., the model presupposes some age-related decline in organ system function in order to rationalize the evolution of aging.

OK, so suppose that the old remain fertile, but eliminate themselves to avoid competition with their own offspring; reproductive senescence then evolves later since there’s no positive selection pressure for maintaining reproductive function over the long term. Problem: What’s the point? If both old and young are making copies of the same genes, there’s no fitness advantage in eliminating the old — especially in light of the fact that most of the offspring’s competition would be coming not from their own parents and grandparents but from more distantly related members of the same species. (And in sexual organisms, you are a better copy of your own genes than your offspring, who have only half of your alleles. Far better to stick around and show the kids how it’s done, than ride off into the sunset to clear the path for these dilutions of oneself.)

Group selection of aging is also vulnerable to “defectors” — mutants who take advantage of the situation to spread their own selfish genes. Suppose that there is some species-level advantage to aging, such that it emerges as a positively selected trait. As organisms age, they actively decrease their own viability in such a way that they have an increased mortality. The species benefits (somehow) at the cost of the individual fitness of these “cooperators.” But then along comes a defector mutant, who doesn’t age and continues to reproduce while the cooperators are pushing up the daisies. Unless the species-level advantage is overwhelming, it’s clear that the defector trait will spread within the population.

Ultimately, then, the reason why group selection models don’t satisfactorily explain the evolution of aging is that it’s hard to imagine a scenario in which a species-level advantage conferred by aging could outweigh the organism-level advantage conferred by not aging.

Such a scenario might now have been imagined. Mitteldorf and Pepper postulate that senescence could have evolved in order to prevent the spread of disease epidemics in populations:

Senescence as an adaptation to limit the spread of disease

Population density is a robust measure of fitness. But, paradoxically, the risk of lethal epidemics which can wipe out an entire population rises steeply with population density. We explore an evolutionary dynamic that pins population density at a threshold level, above which the transmissibility of disease rises to unacceptable levels. Population density can be held in check by general increases in mortality, by decreased fertility, or by senescence. We model each of these, and simulate selection among them. In our results, senescence is robustly selected over the other two mechanisms, and we argue that this faithfully mirrors the action of natural selection. This picture constitutes a mechanism by which senescence may be selected as a population-level adaptation in its own right, without mutational load or pleiotropy. The mechanism closely parallels the ‘Red Queen hypothesis’, which is widely regarded as a viable explanation for the evolution of sex.

OK, so, how might this work?

Epidemiology is, by definition, a population-level issue, and there’s already precedent for selection pressure based on disease susceptibility guiding evolution at the species level (e.g., the diversity of major histocompatibility loci).

The trick is to get the pressures at the individual and group levels to point in the same direction: If I (an organism) am more susceptible than average to a given disease, and that susceptibility has a genetic component, then my closest relatives (who share most of my genes) are likelier than the general population to be susceptible as well. Therefore, my continued existence poses a risk for my progeny, because I represent one more potential host for a pathogen that might infect them – potentially killing us all and ending the line altogether. One way to deal with that problem is to eliminate hosts, and the authors’ model shows that senescence is a reasonable way to achieve that end.

ResearchBlogging.orgMitteldorf, J., & Pepper, J. (2009). Senescence as an adaptation to limit the spread of disease Journal of Theoretical Biology DOI: 10.1016/j.jtbi.2009.05.013



  1. Then by that logic should we expect organisms with unusually long lifespans for body size to be genetically more resistant to disease? And should the key to aging be correlated with disease resistance?

    This could be ignorance on my part, but can someone give me an example of a lethal epidemic wiping out an entire population?

  2. Thanks for the reference. I suppose that works, but not in terms of population growth. Thinking of it now, I should just download the Mittledorf paper and see what they reference.

    What is interesting about your reference is that it ties in to global warming. An interesting issue in light of the previous post on this blog on how ideas can be spread in science. I guess fads dominate science for better or worse.

  3. “Population density can be held in check by general increases in mortality, by decreased fertility, or by senescence. ”

    Or by migration. Individuals having agoraphobia and trying to move from populated areas to less populated areas would generate a tendency to the species to mantain a uniform density and it’s a trait that can be found in animals. This model wouldn’t work in a island simulation but would in a continent style simulation where the barriers for an expansive species are surmontable eco barriers and competition.

    It would be interesting to add this to the simulation and run it again.

  4. Diseases and parasites also adapt to a population over time, so the aged members may well be more susceptible, thus present even more of a risk per capita, than the young.

    Does this imply that populations living in conditions that favor epidemics will be shorter-lived than populations living in less risky conditions? Would crowding, food surpluses, lack of competition, or absence of predators result in shorter lifespans? I’d be interested to see whether social species have shorter lives than rare, loner species, to a degree not explained by the increased difficulty in finding mates.

  5. As a point of clarification: Despite the way I framed it in the post, I don’t think that an epidemic would have to wipe out an entire lineage in order to provide selection pressure. As long as lineages with less senescence were differentially vulnerable to disease, the pressure could play a role in selecting for the maintenance of senescence.

    I like Alexandre’s point about migration — it implies a prediction that organisms that don’t stay near their own relatives should be longer-lived, all things being equal.

  6. I may be misreading this, but are you presuming this to be a form of group selection? As I see it the benefit is directly given to the related kin, so unless you consider kin-selection to be group selection this still can be looked at as a form of individual selection.

  7. @Dave: In the model, there’s a benefit to the species as a whole from the evolution of senescence. The kin selection issue prevents defection. So, is this group selection or individual selection? It’s sort of a hybrid case. Your point is well taken, however.

    @Peter: It’s not my theory; it’s the theory of the authors of a paper that I blogged about. That said: Humans haven’t been living in their modern population distributions for long enough to evolve as a result of those distributions. Macau certainly wasn’t dense prior to 100 years ago.

  8. Shouldn’t we see, according to this hypothesis, a tendence to shorter lifespans, particularily in animals prone to live in groups & face pandemics?

  9. I’d want to see the full paper, but my guess is that the model is oversimplified. I bet it doesn’t adequately take into account other factors associated with large population densities (resource depletion, predator population, etc.).

    Given the power of Natural Selection I would expect senescence to be rapid and triggered by a combination of density dependent factors and reproduction if this hypothesis was accurate, since senescence would only be advantageous at high population densities and disadvantageous at low densities. Actually, sound a bit like salmon…

    Hmm, I’ll get the article…

  10. Aging itself is reaching epidemic proportions. It’s not natural to have so many old people.

    The money is on how young people will live in rapidly aging societies.

Comments are closed.