In chapter 1, we covered some of the science behind life extension, particularly possible methods of life extension, the basic processes of aging, and how long we might live. That was enough to get the book going, and I didn’t want to delay the discussion of ethics. However, some readers may want to know more about the science behind all this. There is good reason for that: knowing more will deepen your understanding of life extension, and the science is fascinating. This appendix does not repeat the material presented in chapter 1; it complements that material.
I defined life extension as slowing, halting, or reversing some or all aspects of aging, but what is aging itself? First, consider several things that aging is not. Aging is not merely the passage of time: two organisms might be the same chronological age, but one might have aged more than the other (it looks and feels older). Moreover, aging is not merely the visible physical changes we associate with old age, such as gray hair, wrinkles, fading hearing and failing vision, or loss of bone tissue, for different species and different individuals may display different outward signs of aging even when their life spans are the same and they have aged the same amount. Nor should aging be identified with whatever cellular- and molecular-level processes lie behind increasing frailty over time, for it’s conceivable that two species both exhibit aging but have different cellular- and molecular-level processes of aging. We have a clear intuitive sense of aging (we know it when we see it), but for a more precise definition, we need to consider how geroscientists define aging. Their definitions tend to fall into two groups.
Some geroscientists define aging in terms of universality: unlike disease, aging is found in all members of a species. Bernard Strehler, for example, says that aging has four characteristics: it is universal (found in all members of the species), intrinsic (inherent to the organism and not a response to environmental factors), progressive (it gets worse over time), and deleterious. Some geroscientists add two more characteristics: aging is also irreversible and genetically modified.1 The universality element is the heart of these definitions, however, for some diseases are intrinsic, progressive, deleterious, irreversible, and genetically modified, but none of them is universal.
Most geroscientists define aging in terms of increasing risk of failure.2 According to this definition, aging is simply a state of affairs where, over time, organs and other systems in the body are increasingly likely to fail. According to the risk of failure definition, the very fact that they’re more likely to fail at 30 than they were at 20, for example, is aging. This definition began with Benjamin Gompertz, the English actuary who discovered the “Gompertz curve”: after age 30, the odds of dying double every 7 to 10 years (depending on the country). John Maynard Smith has this in mind when he defines aging as “a progressive, generalized impairment of function resulting in an increasing probability of death.”3 The increasing probability of death is due to failures in function that result in sickness, disease, frailty, and/or reduced speed or endurance. Does this mean that all failures of biological function are part of aging? If my appendix fails at age 25, have I thereby aged? Presumably not; the failures might be associated with aging (glaucoma, for example), but aging itself is whatever makes the odds of such failure greater over time, even if we think of the failures themselves as part of aging.
The universality and risk of failure definitions might be coextensive in what they pick out; it may turn out that whatever causes increasing risk of failure in a species is also universal to all members of a species, and vice versa. That said, I favor the risk of failure definition, for it has some operational virtues. First, we can use it to date the beginning of aging in an individual: an individual starts to age when the risk of failure begins to increase. (Steven Austad suggests that human aging begins at the age where the odds of death are at a minimum: age 11.4) Second, we can also use it to measure the rate of aging—just measure the rate at which failure is increasing. Third, the risk of failure definition enables us to entertain the possibility that some aspects or processes of aging might not be universal—that is, the possibility that some individuals might suffer increasing risk of failure in certain processes or systems and others might not. (My intuition is that this is aging even if it’s not universal.) Finally, the risk of failure definition doesn’t require a sharp distinction between aging and age-related disease. We do want to distinguish diseases that aren’t part of aging, but that doesn’t require claiming (as universality definitions do) that no age-related diseases are part of aging; the risk of failure definition allows us to entertain that possibility too. I am not claiming that they are part of aging, only that we should not settle the question by defining them out of the concept.
So life extension is slowing, halting, or reversing increases in the risk of failure in an organism.
The first thing to understand about life extension is that it will be nothing like the gains in life expectancy we’ve seen over the last 150 years. Those gains didn’t come from slowing aging. They came from eliminating things that kill us before we reach our natural life span.
The maximum natural human life span is roughly 120 years. It hasn’t changed in 100,000 years, but few of us live much past 80. It could be worse; the archeological record indicates that for most of human history, hardly anyone lived past 45. Life expectancies increased partly because societies achieved a more stable food supply through agriculture, partly because of improved sanitation and living conditions, and partly through eliminating infectious and parasitic diseases. As a result, life expectancy at birth in developed countries is now in the upper 70s or higher. People often think this means that adult life spans are longer than they were a few centuries ago, but that’s a mistake; between 1860 and 1960, life expectancy at birth in the United States increased by 31 years, but the life expectancy for an adult who reached the age of 60 increased by only 1 year during that century. In other words, a century ago, those who made it to 60 lived roughly as long as we do, but the odds of making it to 60 were much worse. A high percentage of the increase in life expectancy comes from reducing infant and child mortality. Note that none of this involved slowing the rate at which people age.
This increase in life expectancy is known as the epidemiological transition. Thanks to this, people in developed countries rarely suffer or die from infectious diseases. Instead, they suffer and die from age-related chronic and degenerative diseases, such as cancer, heart disease, stroke, and diabetes. However, unlike what happened when we eliminated infectious diseases, eliminating these diseases would not produce big gains in life expectancy. Eliminating all cancer, heart disease, stroke, and diabetes would increase our average life span from 82 years to 96 years—an increase of 17 percent.5
We’re nearing the end of what can be done to extend the human life span by eliminating disease. From here on, the only way to dramatically increase life expectancy is to slow the rate at which we age.
We’re so familiar with aging that we take it for granted. However, aging is very puzzling for several reasons. First, it’s a puzzle from the standpoint of evolutionary theory. The longer an organism lives, the more offspring it can have. Natural selection favors organisms that produce more offspring. Given that organisms can age more slowly (as the experiments mentioned above indicate), we would expect evolution to select for organisms that age very slowly or not at all and therefore have more offspring. However, evolution has not done this.
Second, organisms develop from single-cell fertilized eggs into vastly complex living things. Creating a complex organism is more difficult than maintaining one. If creating an organism is harder than maintaining it, we should expect that organisms would maintain themselves indefinitely and never age. After all, they’ve already performed the more difficult task. However, they don’t.
Third, geroscientists have recently discovered that we gradually cease to age in our mid-90s.6 Aging is an increase in the risk of failure for an organism or its components; the older you are, the more likely you are to have a heart attack, for example. Our risk of failure starts to increase during late youth and gets worse as we get older. However, it levels off around age 95. This is called the late-life mortality plateau: if you live to around 95, you remain aged and grow older, but you do not age any further. Of course, your risk of failure is so high by then that you eventually die of some age-related cause. Even so, our bodies naturally stop aging if we just live long enough.
Fourth, aging is puzzling because some organisms never age, or they do it too slowly for us to notice. Rockfish, the bristlecone pine, lobsters, sturgeons, the microscopic hydra, and some sharks do not age at all—accidents, disease, and predators finish them off. Male flounders age but female flounders don’t (they just keep getting bigger). Some colonies of corals are more than 20,000 years old.7 Bacteria don’t age; they divide into daughter cells—an ending that’s not quite a death. Other species age but nonetheless regenerate limbs and other parts of their bodies; starfish, some reptiles, and all plants can do this. Even we do this to some extent, for most cells in the human body are less than 10 years old and most will be replaced several times during your life. Within each cell, molecules are constantly being broken down and replaced, even in cells that are not themselves replaced, thereby rebuilding the cell from the inside out. Our germline cells—the cells that produce eggs and sperm—don’t age at all.
So why do we age? Aging is puzzling from the standpoint of evolutionary theory, and the answer to that puzzle lies in evolutionary theory. There are four evolutionary theories of aging that claim to solve that puzzle.
The first evolutionary explanation of aging is the mutational accumulation theory associated with Peter Medawar.8 Some genetic defects appear early in the life of an organism and others appear late in its life. This theory builds on the fact that if an organism has a genetic defect that appears early in its life, the defect will make the organism vulnerable in early life, before it can reproduce, or at least before it reproduces as much as it’s capable of doing. If that happens, the genetic defect is less likely to be passed on to offspring—or at least not to very many of them—for the organism will die before it has very many offspring. Thus, natural selection works against genetic defects that appear early in life.
However, things work out differently with genetic defects that appear later in life. A defect that appears later in life is likely to be passed on to a larger number of offspring, for the organism will reproduce before the defect weakens it. If the defect appears late enough, the organism is likely to die from predators, disease, or resource shortages before the defect materializes. Thus, defects that don’t appear until that time or later are unlikely to be weeded out by natural selection, for those defects don’t affect reproductive success. Finally, the later in life the defect appears, the less likely it is that natural selection will weed it out of the population. For example, progeria (a genetic disease that mimics aging and kills by the time of adolescence) occurs only through random mutation—it’s never inherited—because its victims die before they can reproduce. (Because it’s random, it’s also quite rare.) However, Huntington’s disease doesn’t kill its victims until they’re in their 40s and have had a chance to reproduce. Natural selection doesn’t weed this defect out of populations, and it therefore runs in families.
The result is that harmful genetic mutations that occur only late in life will accumulate—hence the name “mutational accumulation theory.” According to this theory, the phenotypic changes we call aging are caused by genetic defects that did not, before the rise of agriculture and civilization, affect individual survival, for humans died of something else first. Therefore, those genetic defects were passed on to offspring. However, under civilized conditions, most people live long enough for those genetic defects to manifest themselves, and thus we age.9
However, George C. Williams noticed some problems with this theory. First, animals in the wild do sometimes die from senescence. They may not appear old, but senescence is gradual in most species, and even a bit of aging may slow an animal down enough for a predator to catch up. Another problem is that some of the genes involved in senescence are not random mutations but have been around for millions of years; baker’s yeast, nematode worms, fruit flies, and mice all share some genes involved in aging. These species do not all die at around the same age in their natural habitat, so the age when they are statistically likely to die from predators or disease does not obviously explain how they all came to share these genes.
Williams proposed the second major evolutionary theory of aging: the antagonistic pleiotropy theory.10 Genes exhibit “antagonistic pleiotropy” when they have both beneficial and harmful effects. Williams surmised that some genes might have a beneficial effect early in life and a harmful effect later in life. According to his theory, the genes involved in aging have beneficial effects early in life, while late-life effects result in the changes we call aging.11 For example, the genes responsible for testosterone have a beneficial effect early in life, for testosterone promotes reproduction. However, later in life, testosterone accelerates deterioration in arterial walls, suppresses the immune system, and helps cause prostate cancer—changes related to aging.
The antagonistic pleiotropy theory has problems too. First, some genes involved in aging do not have beneficial effects in life. Second, as for those genes that do have beneficial effects early in life and harmful effects later in life, random mutations could produce modifications that eliminate the later harmful effect without losing the early beneficial effect, so one would expect the harmful later effects to disappear from the population as individuals who lacked them outbred their competitors.
The third theory is Thomas Kirkwood’s disposable soma theory. This theory starts from the premise that an organism has a limited amount of energy to allocate between maintaining its tissues against the basic processes of aging and everything else: courtship, breeding, hunting, food gathering, fighting, and fleeing. Animals whose tissues renew themselves to maintain a perpetually youthful body are using energy that could otherwise be used for reproduction, fighting, and so on. Because natural selection favors species that reproduce more, it favors species whose individuals devote just enough energy to repairing their bodies to live long enough to reproduce as much as those individuals can in their environment—and no more. Tom Kirkwood calls this the “disposable soma” theory because the soma (a “soma” is a body) is a disposable vehicle for combining and distributing the germline cells (eggs and sperm)—somewhat like using up a booster rocket to get a capsule into space. Once that task is accomplished, there is no further need for the soma.12
The disposable soma theory faces obstacles as well. First, there is the caloric restriction effect: animals live longer on low-calorie diets with sufficient nutrition. Although nutrition may not be reduced, metabolic energy is reduced, and that should leave less metabolic energy for repair—yet the animals live longer, not shorter lives. Second, females need more metabolic energy for reproduction than males (for gestation and other reasons), yet males in many species live roughly as long as females.
The problems discussed above do not necessarily refute these three theories, and they each have many supporters. They are also compatible; it may turn out that some genes exhibit antagonistic pleiotropy, while others are late-life defects that natural selection does not weed out, and that organisms evolved to conserve metabolic energy for purposes other than needless longevity.
The fourth evolutionary theory of aging is the programmed aging theory. This is the earliest of our four theories.13 In 1889, August Weismann suggested that aging evolved in order to make room for new generations; if there was no room for new generations, then a species would die out. However, this theory has fallen out of favor, for it faces several serious objections. First, animals in captivity live longer and exhibit aging, but in the wild, they don’t live long enough to visibly age. If aging is triggered by a genetic program of some kind, we would expect all animals to die when the program kicks in, regardless of where they live. Second, in some animals (including us) there’s a late-life mortality plateau. Humans stop aging in their mid-90s. This suggests that the aging program shuts off after a certain age, but that means that those survivors are not being cleared away to make room for new generations. Third, random mutations could produce individuals who either lack the program or have some way of shutting it down and living longer. If they live longer, they will have more offspring and spread their genes through the population, eventually eliminating the death program from the gene pool. However, that seems not to happen.
Recently the programmed aging theory has been revived (though it remains a minority view) using recent developments in evolutionary theory concerning group selection and evolvability.14 Group selection theory says that evolution sometimes selects for traits that benefit a group but harm individuals, such as altruism. For example, a population that regulates its own population is less susceptible to mass famine or mass epidemic, for it won’t overpopulate its habitat as quickly, thus leaving more resources for times of famine or epidemic (this is group selection). Evolvability theory says that evolution sometimes selects for traits that are disadvantageous to an individual but offset that disadvantage by facilitating the individual’s ability to evolve. For example, an animal that’s older and more experienced but less endowed with intelligence or immunities may survive better than a younger, less experienced animal that has more of those qualities. However, if the older animals in that population die off, then the younger ones acquire experience while retaining their intelligence or immunity. This produces a more fit population overall—thus populations programmed to age may be reproductively more fit. Similarly, if individuals in a population are programmed to age after a certain span of time, the population will evolve faster because there is more room for new generations and hence new mutations. The upshot is that evolution selects for aging, not as a by-product or accident, but because it’s advantageous to a population.
The programmed aging theory predicts that there are genetic programs that induce aging. There is some evidence that such programs exist. First, there is apoptosis—programs that kill individual cells that are infected, cancerous, or need to be removed during development (like the webbing between fingers and toes during fetal development). Apoptosis occurs more often late in life, also killing healthy cells. Second, there is replicative senescence—some cells can divide only a limited number of times before their telomeres (genetic caps on the ends of DNA, like the caps on your shoelaces) are too short to allow further division, and then they die because the cell cannot replicate its chromosomes without cutting into its genes. When telomeres get short enough that these cells stop dividing, they seem somehow to alter their gene expression to permit oxidative damage and inhibit cell repair.
These four theories of aging have different implications for the possibility of life extension. The mutational accumulation theory suggests that the mechanisms of aging are quite various and that slowing or halting aging cannot be achieved by tweaking our supposed natural defenses against aging, for it doesn’t imply that such defenses exist. It implies that there is no control panel for aging—just a collection of maladies that require a collection of remedies. (By “control panel” I mean a limited set of genetic mechanisms that control all aspects of aging.) That doesn’t make life extension impossible, just more difficult to achieve. The antagonistic pleiotropy theory faces that problem too, for it says that aging results from a variety of genes with beneficial effects early in life and harmful effects later on.
The prospects for life extension look brighter on the disposable soma theory and the programmed aging theory. According to the disposable soma theory, aging results from a failure of the maintenance systems we already have. It’s possible to slow or halt aging because the body already does, at least during youth, and to a lesser extent later on. Moreover, aging occurs all over the body at roughly the same time in multiple organs and tissues, which suggests one set of maintenance systems. This is also true of the programmed aging theory, except that instead of systems that slow or halt aging, we have systems that cause it—again, all over the body at roughly the same time, which suggests one set of aging systems. On both of these theories, there may be fewer basic processes of aging and thus a control panel for aging.
Of course, it may also turn out that all four theories contain part of the truth and that there is a control panel for some aspects of aging but not for others, thereby complicating the search for life extension.