Comparative Biology of Aging

After witnessing the amazing diversity of aging phenotypes and lifespans found in Nature, a key question is whether there are any trends in the way animal age. In this essay, I briefly examine the main factors associated with longevity in animals. Some of the trends observed can inform our knowledge of aging mechanisms. Recent advances in genomics and how these may contribute to unravel the biodiversity of aging are also discussed below.

Keywords: ageing, allometric scaling, allometry of life, evolutionary genomics, life span

"The reasons for some animals being long-lived and others short-lived, and, in a word, causes of the length and brevity of life call for investigation."

Aristotle, 350 BC

Why cannot a mouse live more than 5 years while humans and whales can live over 100? Why do some animals appear not to suffer from aging? In the end, why do different species age at different paces? Many researchers have asked these questions (Austad, 1997a & 2005; Warner et al., 2002; de Magalhaes, 2003). So far, and even though this topic will be discussed in other essays, including at the genetic level, the answer has eluded us. Nonetheless, there some general trends have been reported and some factors shown to correlate with maximum lifespan.

As mentioned before, quantifying aging is a difficult, controversial task. Quantifying the rate of aging for a given species can be done through the MRDT, but there are some caveats: MRDT calculations are available for only a fraction of species and, as shown before, are not perfect estimates of rate of aging. Consequently, most researchers use maximum lifespan (tmax) as an estimate of rate of aging. It has been argued that tmax represents the genetic potential for longevity of each species and is related to a species' rate of aging (Cutler, 1979; Allman et al., 1993; Finch and Pike, 1996). Even though there are potential problems in using tmax to estimate rate of aging, such as biases caused by different in sample sizes, tmax remains the best and most widely available measurement to quantify rate of aging (de Magalhaes, 2006).

One factor that correlates with maximum lifespan is body size. Clearly, the typical adult body mass for a species correlates with tmax (Fig. 1). In other words, larger animals live, on average, longer than smaller animals, as shown and debated by a large number of authors (Calder, 1984; Schmidt-Nielsen, 1984; Promislow, 1993; Austad, 2005; de Magalhaes et al., 2007a). The logarithmic relation between tmax and body mass (M), also called the allometry of lifespan, has been the subject of intense scrutiny. From Figure 1, we can obtain the equation: tmax = 5.58M0.146 with r2 = 0.340. The squared Pearson correlation coefficient (r) suggests that body mass explains 58% of the variation in tmax. Clearly, there are many exceptions to this correlation. One exception are bats that live a lot longer than expected for their body size (Austad and Fischer, 1991; Austad, 2005; de Magalhaes et al., 2007a). Likewise, when compared to mammals, birds live longer than expected for their body size.

Body mass versus maximum lifespan

Figure 1: Correlation between maximum lifespan (tmax) and typical adult body mass (M) using all species (n = 1,701) present in AnAge build 8. Plotted on a logarithmic scale.

At present, the simplest and most likely explanation for the allometry of lifespan is related to ecological constraints: smaller animals tend to be more prone to predation and thus are expected to have higher extrinsic mortality rates, a shorter tmax, and a faster aging process--as debated ahead in more detail. For example, the ability to fly gives most birds and bats the capacity to evade predators. Consequently, it seems that body mass is a determinant of ecological opportunities and habitat that impacts on mortality, which consequently influences the evolution of longevity and aging (Stearns, 1992). To date, there is no evidence to suggest that some unknown physiological affects aging in a way proportional to body mass.

Experimentally, the impact of body mass on tmax is relevant because it can bias comparative studies of aging (Promislow, 1993; Speakman, 2005). Researchers trying to identify factors that correlate with tmax must eliminate the effects of body mass from their calculations, which can be done with certain statistical procedures. In fact, body mass appears to correlate with many life history events besides maximum lifespan: gestation period, time to maturity, etc. Therefore, researchers studying whether a given factor correlates with tmax or not must play close attention to the impact of body size. As I discuss in the context of theories of aging, this has not always been done, however, sometimes resulting in erroneous interpretations of experimental results.

Brain mass also correlates with tmax, even after correcting for the biases caused by body mass. This is particularly true in primates (Allman et al., 1993). The way brain mass appears to be a better predictor of longevity than body mass is probably due to less variation in brain mass (Lindstedt and Calder, 1981). Therefore, even though it can be argued that this relationship shows the influence of the brain on longevity, it does not prove that the causes of aging are located in the brain. In fact, the size of other organs also correlate with tmax, in some cases more strongly than brain size (Austad and Fischer, 1992). Besides, ecological explanations are also possible in that maybe animals with bigger brains are better at escaping predators for a number of reasons.

Even though bigger species tend to be longer-lived than smaller ones, it is interesting to note that there are a number of cases in which smaller animals within a given species live longer in captivity. These include mice, rats, horses, and dogs (Miller, 1999; Miller et al., 2002a; Rollo, 2002). For example, it is well-known that smaller breeds of dogs live longer. Interestingly, it has been argued that "little people" may also be longer-lived (Krzisnik et al., 1999). Therefore, it appears that while on one hand bigger species tend to be longer-lived, within a given species smaller individuals--in protected environments--tend to live longer. The possible physiological and genetic reasons for the latter phenomenon and implications for our understanding of aging are debated in another essay.

Another relationship long studied in gerontology is Kleiber's rule that relates maximum lifespan with metabolic rate (Kleiber, 1975; Gosden, 1996, pp. 103-110). (Kleiber's rule actually originates in a theory of aging called the "rate of living theory," which is discussed in more detail elsewhere.) It can be argued, for instance, that reptilians and amphibians live longer because they have decreased metabolic rates since they are cold-blooded animals. Similarly, if the metabolic rate, the rate at which reactions occur in cells is higher in, for instance, mice than in humans then maybe that is why mice live less than humans (Prinzinger, 2005).

Despite its intuitive nature, there is no evidence that metabolic rates influence aging in endotherms like birds and mammals. First of all, there are gross exceptions: bats and birds live longer than what would be expected for their metabolic rates. In addition, marsupials live less than eutherians and yet have lower body temperatures, which implies a lower metabolic rate (Austad, 1997a, pp. 88-90). Another problem is related to body size. Metabolic rates are often estimated by measuring oxygen consumption at rest. Clearly, an elephant will breath in more oxygen than a mouse, so it is necessary to correct for body mass. Failure to do so will result in oxygen consumption being associated with tmax incorrectly--i.e., due to its relation to body mass which in turn correlates with tmax. When the effect of body mass is correctly eliminated from metabolic rates metabolic rates do not appear to correlate with tmax. In fact, recent results suggest that metabolic rates are not associated with tmax in mammals or birds after correcting for the effects of body mass using the most state-of-the-art statistical methods (de Magalhaes et al., 2007a). The exact methodology of these calculations can be attacked--e.g., because to the way metabolic rates are corrected for body mass or even the way tmax records are obtained. Nonetheless, there are no results in which metabolic rates are correctly adjusted for body mass that show a correlation between metabolic rates and maximum lifespan in mammals or birds. Kleiber's rule is thus mostly discarded now.

Partly related to metabolic rates, a point of debate is whether hibernating species live longer than non-hibernating species. So far the results are mixed, but some results suggest hibernating animals may live longer (see, for instance, Lyman et al., 1981; Brunet-Rossinni and Austad, 2004; Turbill et al., 2011), which could suggest that a period of metabolic torpor could increase lifespan. On the other hand, it can be argued that spending a fraction of the year in hiding, during which time mortality is presumably low, contributes to the observed longer lifespan in hibernating animals.

Growth and development are two other factors that correlate with tmax. Independently of body mass, age at sexual maturity correlates with average and maximum adult lifespan in many taxa, including in mammals (Charnov, 1993; Prothero, 1993; de Magalhaes et al., 2007a). In other words, the longer it takes for a given mammalian species to reach sexual maturity, the longer it can live afterwards. There are some exceptions, however, such as the male Anthechinus which is mentioned elsewhere. One hypothesis is that there is a mechanistic link between pace of development and pace of aging, as discussed in another essay. It is also worth mentioning that each organism's body-plan is largely determined by its genetic program, and the body-plan can have a powerful influence on longevity, as shown by aphagy in some insects or the semelparity of species like the salmon. Different species could well be influenced by development in different ways: the relation (adult phase)/(total lifespan) shows a wide variation, which is in accordance with the several aging phenotypes found in nature. So development and its consequential body-plan can influence aging to different degrees. The body-plan of mammals, for instance, may place indirect constraints on adult life but this could be regarded as a by-product of development. That said, age at maturity correlates strongly with tmax in mammals which hints that common regulatory mechanisms could be involved (de Magalhaes et al., 2007a). Though not as strongly, growth rates also correlate negatively with tmax; in other words, species that grow slower tend to live longer (de Magalhaes et al., 2007a). Likewise, growth rates correlate negatively with demographic rate of aging--not MRDT but a similar parameter estimated from the Weibull model (Ricklefs, 2010). On the other hand, for evolutionary reasons, development can be timed similarly to aging even if the relation between development and aging in mammals is indirect and minimal (Miller, 1999). Therefore, the causes for the relationship between developmental time and longevity remain a subject of debate, though it is clear that there is a strong correlation between them.

Recently, the declining costs of DNA sequencing have led to increasingly more powerful approaches in comparative genomics (reviewed in de Magalhaes et al., 2010). Sequencing the genome of an organism is no longer a large-scale endeavour and the genomes of hundreds of species are currently being sequenced. This is exemplified in the sequencing of the long-lived naked mole-rat (Kim et al., 2011). The transcriptome (i.e., RNA) can also be sequenced in a cost-effective fashion to obtain, for example, measures of gene expression levels. Again using the naked mole-rat as an example, transcriptome sequencing revealed that genes associated with oxidoreduction and mitochondria were expressed at higher levels in naked mole-rats when compared to mice, which may contribute to the naked mole-rat's longevity (Yu et al., 2011). The availability of multiple mammalian genomes also opens the door to try to identify gene features associated with longevity. For example, methionine residues in mitochondrially encoded proteins appear to be enriched in short-lived species (Aledo et al., 2011) and cysteine residues appear to be depleted (Moosmann and Behl, 2008). Comparisons between nuclear genomes across species with different lifespans can also focus on identifying genes with patterns of evolution associated with longevity (de Magalhaes and Church, 2007). One genome-wide scan for genes associated with the evolution of longevity in mammals found evidence that proteins involved in protein degradation, a process associated with aging, are under selection in lineages where longevity increased (Li and de Magalhaes, 2013). Given the explosion of genomic data, these approaches are bound to become more powerful and reveal specific genes and patterns associated with longevity. To facilitate comparative studies of aging, including in genomics, our lab has developed the AnAge database which features thousands of longevity records for animals (reviewed in de Magalhaes et al., 2009b). As discussed elsewhere, I think this shift from comparing physiological traits into digital biology will have a major impact in furthering our knowledge of mechanisms of aging.

Up to the Gerontology Information

Back to

Thank you for visiting my website. Please feel free to contact me if you have any questions, ideas, comments or suggestions.

Copyright © 1997 - 2001, 2004, 2005, 2007, 2008, 2010, 2012, 2013 by João Pedro de Magalhães. All rights reserved.