Harvard University
INDEX
- Welcome and Opening Remarks
- Session 1: Adding Years to Life: Current Knowledge and Future Prospects
- Session 2: Duration of Life: Is There a Biological Warranty Period?
- Session 3: Prescription Stimulant Use in American Children: Ethical Issues
- Session 4: What's Wrong with Enhancement: Discussion of Paper Prepared by Michael J. Sandel, D. Phil.
Welcome and Opening Remarks
CHAIRMAN KASS: Good morning everybody. Welcome to Council Members, to guest presenters and to members of the public. I would like to recognize the presence of Dean Clancy, the designated federal officer in whose presence this meeting may officially begin.
I would also like to extend a special welcome to the members of Professor Sandel's class from Harvard University who have come down to witness how this conversation takes place inside the beltway. They've been studying with Michael this quarter, and we're delighted to have you here.
This meeting will be the third meeting on our project called either "Enhancement" or "Beyond Therapy", in which we are exploring the possible uses of new biomedical technologies beyond the treatment of individuals with known diseases and disabilities, uses either for personal enhancement, the satisfaction of client desires, or for social and behavioral control. And this meeting from beginning to end will explore technologies that might affect the native in-born capacities of human beings through the uses of genetic and genomic knowledge, that's tomorrow, that might affect human behavior above and beyond the treatment of disease through the use of stimulant drugs this afternoon, and this morning, technologies that might push the temporal boundaries and trajectory of a natural human life span through research on the biology of aging.
The first two sessions this morning devoted to aging research will explore first, the question of whether we can add years to life and exploration of the current research and future prospects, and the question of the duration of human life, whether there is such a thing as a biological warranty period.
If I might, since this is a topic dear to my heart, if you might indulge me, I would like to read a couple of pages from 20 years ago, when I knew more about this subject than I now do. But this is an introduction. Actually, by the way, when I was the Staff of NAS Committee on Life Sciences and Social Policy, which is almost 30 years ago, one of the chapters of our report was on the retardation of aging as the futuristic possibility that nevertheless raised large questions, and this is the introduction of an old essay.
"Why should we die? Why should we, the flower of the living kingdom, lose our youthful bloom and go to seed? Why should we grow old in body and in mind, losing our various powers, first gradually, then all together in death? Until now, the answer has been simple, we should because we must. Aging, decay and death have been inevitable as necessary for us as for other animals and plants from whom we are distinguished in this matter only by awareness of this necessity. We know that that we are, as the poet says, like the leaves, the leaves that the wind scatters to the ground.
Recently, this necessity seems to become something of a question thanks to research into the phenomena of aging. Senescence, decay, and even our species-specific life span are now thought to be the result of biological processes that are, at least in part, genetically controlled, open to investigation, and in principle, subject to human intervention and possible control. Slowing the processes of aging could yield powers to retard senescence, to preserve youthfulness, and to prolong life greatly, perhaps indefinitely. Should these powers become available, whether to wither and why will become questions of the utmost seriousness."
And then I make a series of arguments as to why we should take these up even now, even though these are futuristic matters, and go on to point out that the prolongation of healthy and vigorous life, and ultimately perhaps even a victory of mortality was, perhaps, the central goal and meaning of the modern scientific project as articulated by its founders, men such as Bacon and Descartes.
Bacon it was who first called humankind to the conquest of nature for the relief of man's estate, and there's ample suggestion in Bacon's writings that he regarded mortality itself as that part of man's estate from which he most needs relief. Bacon himself engaged in immortality research, and may well have been its first martyr, sacrificing his life on the altar of longevity. He apparently contracted his fatal illness while performing freezing experiments on a chicken.
Descartes in the famous passage in Part 6 of the Discourse of Method, where he rejects the speculative philosophy of his predecessors in favor of a practical philosophy that would render ourselves as masters and possessors of nature, goes on to talk about the benefits of this new power, amongst which he says that we could be free of an infinitude of maladies, both of body and mind, and even also possibly of the infirmities of age if we had sufficient knowledge of their causes, and of all the remedies with which nature has provided us.
This is an old story, and it's been a dream not just of magicians and sorcerers, but even of the great founders of modern science. The success of the past century, which increased the average life expectancy at birth from 47 to 77 is a success that cannot be repeated, since that increase was largely due to the conquest of childhood diseases, sanitation and the like. But further increases in the potential human life span, as it's been pointed out in our readings, could not come from curing the specific diseases that now afflict us even in our old age, and the debilities of old age, the weakness, the brittleness, the decline in bodily and mental powers remain. Retardation of aging through understanding of the basic processes holds the key, if there is one, both to adding life to years, and of adding years to life.
This Council takes this up, not because we've been taken in by cryopreservation or by the vast arrays of creams and elixirs that are now being sold to a gullible population of retirees and aging baby-boomers, but because of the exciting new developments in the field of aging biology itself. Aging research might turn out to be the ultimate enhancer. Who knows? But in order to separate fact from fiction and to help us understand where this field is going and what it means, we're extremely fortunate to have two of the leading researchers and scholars in this area, people whose research is not only first rate, but who have taken pains to try to bring the meaning of this work to a larger public.
For the first session, we'll hear from Steven Austad, who's Professor in the Department of Biological Sciences at the University of Idaho, and the author of a book, Why We Age: What Science Is Discovering About The Body's Journey Through Life. And in the second session, to my left, Jay Olshansky who is the Professor in the School of Public Health, in the Division of Epidemiology and Biostatistics at the University of Illinois-Chicago, also connected with the Center on Aging at the University of Chicago, and a Fellow at the London School of Hygiene and Tropical Medicine. And with his colleague, Bruce Carnes, the author of The Quest For Immortality: Science At The Frontiers of Aging.
The procedure will be as usual. We'll have presentations from Professor Austad and discussion, we'll take a break, and we'll hear from Professor Olshansky. The floor is your's, and thank you very much both for joining us this morning.
SESSION 1: ADDING YEARS TO LIFE: CURRENT KNOWLEDGE AND FUTURE PROSPECTS
DR. AUSTAD: Thanks for the invitation. The Council suggested six questions that I should try to answer in my presentation, and I'm going to march through those in a fairly straightforward way, but I'd like to start off with a little bit about what I feel is the rationale for the effort to slow human aging. And this is largely taken from the similar reasoning in the paper by Dr. Miller that is included in your briefing booklet.
The goal of I would say most of us in this field is not really the prevention of death, but the preservation of health, and I think from that perspective that goal seems consistent with all of the disease-based biomedical research efforts that we're much more familiar with, the efforts to cure heart disease, Alzheimer's disease, et cetera.
The third point is that if we continue to increase longevity by the sort of disease-based advances, which we've become so good at making over the last century, that we could be facing a major social catastrophe, and I'll just give you one example of that. Neuroscientists and my friends in the Alzheimer's Disease community tell me that approximately 50 percent of people over the age of 85 have some sort of disabling dementia. Therefore, as more and more people, as a larger fraction of the population reaches this age, we could be faced with the possibility of a vastly expanding population of people who need 24 hour a day nursing care.
And the last point is that slowing aging is really a much more effective approach to preserving health, than is the treatment of individual diseases, and I'll give you the rationale for that in this slide here, which shows that these are major causes of death. And you can see that virtually all of them increase exponentially with age. And one of the consequences of the analyses that Jay Olshansky will, no doubt, talk about later, is that curing each of these individual diseases has a surprisingly small impact on life expectancy. But more important, curing one of those diseases does not take care of all of the other disabilities that may be associated with aging, because of other disabilities, such as chronic arthritis, the decline in sensory capacity. These things also increase exponentially in aging, getting rid of one cause at a time, basically leave people who may be alive, but may be very disabled. By slowing down the aging rate, we basically delay the onset and the progression of a whole host of mortal and debilitating diseases.
Now before I get into the questions proper, I would like to talk a little bit about the history of human longevity. From many months that I spent in remote parts of Papua, New Guinea, I got very interested in how long people lived in a state of nature, because when I looked around me there and talked to the people there, the population structure was remarkably young. So I've actually spent a number of years trying to compile the best information possible on how long people lived in the ancient past, and there's really basically two kinds of data.
There is evidence from preserved human remains, and some of this evidence is actually quite compelling; that is, it's based on large samples of well preserved skeletal evidence that's subject to the sorts of forensic analysis that the police often do. And the second is the evidence from the demography of modern hunter/gatherer populations. And I think when you put these two bits of information together, you come up with a relatively persuasive case, because the trends agree so much between these two sorts of evidence. So here are the best two data sets, I believe.
The first data set, Ohio, a thousand years ago, life expectancy was just under 20 years. That's the number in parenthesis there, and you can see, and we'll see a -- I'll show a number of these. These are called "survival curves", and they simply represent the fraction of the population that's still alive at any time. And you can see through this Ohio population of Native Americans, and this is a sample of over 1,300 skeletons, that not only is the mean age at death 19.8, but really there are no deaths over the age of 60.
Now in reality, there were some remains that were estimated to be around 70 years of age. Those don't show up in this analysis, simply because they didn't feel confident enough of their estimates over the age of 55, and so they just lumped all of the estimates, but I'll just tell you the oldest estimate was 70 years of age.
The second bit comes from studies of the modern Yanomami, Native Americans from Southern Venezuela, Northern Brazil, which have been studied by groups at the University of Michigan since the early 1960s, and from interrogations of those people about their ancestors, as well as current demography, we come up with a reasonably similar estimate of a life expectancy of about 17 years. But here you'll notice that the extremes go out past 80 years. Now the actual oldest age reported among these several hundred Yanomami was in the mid-70s. However, due to the demographic modeling of the data, the expectation is that in a large population, some individuals live into their early 80s, so this is basically humans in a state of nature.
And to contrast this by where we've come now, I've put the same two curves up with a curve for males in 1900, and males in the year 2000. And you can see dramatic increases in longevity, more than a tripling of life expectancy, and clearly longer longevity at the extremes. And I think the extremes may be relevant to people's thinking on this issue; that is, I noticed in the briefing booklet that it said that modern humans in the extreme basically don't live any longer than our human ancestors, and there's simply no evidence to support that case. I think that we not only live longer on average, but live dramatically longer in the extremes, and I'll present a little bit of evidence on that in a second.
Now if we look at the bottom graph in this particular -- this shows the probability of dying at any specific age, and the top curve is the curve for women in 1900, and the bottom curve is the curve for women basically today. And you'll notice that the enormous changes in life expectancy are mainly due to enormous changes in the mortality rate at ages prior to the age of 50. However, there are also reductions in the mortality rate straight throughout lifetime for as long as we have data, so even though the big changes were early in life, childhood and early adult mortality, there clearly have been reductions in mortality throughout life.
Now evidence that maximum longevity has increased over what I call archeological and historical time. There are indications from a variety of studies that humans in the state of nature never lived as long as 90, probably not as long as 80 years. Certainly, there have been a handful of modern hunter-gatherer studies and a whole variety of what are known as paleodemographic studies. None suggest anyone living as long as 90.
Prior to the turn of, I guess what's now two centuries ago, there was really no well-authenticated instance of anyone living to the age of 110 years. Now part of that is very likely the consequence of just simply bad documentation, but I think it's arguable that, in fact, it's also attributable to the fact that no one lived that long.
By the turn of the last century, there are approximately 250 well documented to have lived to be 110 years old or older. About 40 of these people are known to be alive now, and this is probably the edge of the wedge, because in places like China, we simply do not have the information. And the maximum longevity now, as of 1977, was Jeanne Calment who lived to be more than 122 years old, something that is statistical outlier even today, but clearly a new phenomenon in human history.
And finally, John Wilmouth, demographer at the University of California, has done quite extensive analyses of the very good Swedish demographic record. It's about 200 years old, and has noted that at least for the last 130 years there's been a steady increase in the maximum longevity, so humans clearly are living longer on average, and at the extreme today than they ever have. I don't believe anyone thinks this is a consequence of anything except a change in the hostility of the environment, improved public health measures, improved medical care. It's not a change in human biology.
So taking that as the introduction, I'll now march through the six questions I was asked to address. The first question was how do modern biologists understand the term "aging", and I've presented a couple of examples here. The one on the top is simply the world record running speeds for the 5,000 meter run. One of the things that you'll note is that not surprisingly, the fastest running speed is among people, as it was 100 years ago, in their 20s, and there's a gradual decline until late in life. And I like to point this out because these master athletes are probably the healthiest and the best conditioned people in their ages that you can find.
The bottom panel shows basically a decline in physiological function in a plethora of traits from nerve conduction velocity, to cardiac index, to maximum breathing capacity. And that list could be multiplied endlessly. The take home message is that aging is the gradual and progressive loss of function over time, beginning in early adulthood. It leads to decreased health and well-being, and an increasing incidence of death, disability I might say, as well as disease. So I think that that definition of aging, which is one that could have easily been made 30 years ago, stands as a reasonable introduction to the topic.
The second question I was asked to address was what happens physiologically as we age, and the basic underlying theme is that there is a generalized decline in what seem to be physiological control mechanisms. Let me give you a handful of examples. Cell population dynamics, we have certain cells in our bodies that are designed to die at appropriate times when they're genetically damaged, for instance. Well-preserved cell population dynamics means that a cell dies when it should.
We also have certain cells that need to replicate on schedule, and well-behaved cell population dynamics will have cells that replicate obediently when they should. However, that control, both in the control of cell replication, the control of cell death gradually deteriorates over time, and so we're led to results of declining repair capacities because our cells may not proliferate as well as they previously did, and also the development of cancers because cells reproduce when they should not.
Certainly, on a cellular level we are beset by huge changes in protein destruction, the proteins that make up the functioning parts of our cell really are designed to turn-over at a certain rate, to die and be rebuilt, be replaced by new proteins. And that rate slows down with aging, both the rate of the destruction of damaged proteins and the rate of replacement by new proteins. There's a generalized loss of hormonal regulation, that is the production of hormones and our body's ability to respond. There is increasing damage to what are known as permanent cells, cells that do not replace themselves over most of the course of the lifetime, and increasing damage to permanent cells is permanent damage. And those are just a few of the changes. I could go on, and on, and on, but I think it makes the point, that virtually everything that can go wrong, gradually does go wrong as we age.
Now are these various phenomena connected? Twenty years ago, I think the unanimous answer would have probably been no, because there's not any obvious connection, except for the case of reduced metabolism. If we sort of take all of the biological processes, that go on in an animal's body, and you slow them down, and you might expect that just because you've slowed them down in a synchronous fashion that they would be synchronous, and I give you an example here, which is how to slow aging in fruit flies. And it's very simple, you put them in the refrigerator. So if you decrease the temperature at which you keep fruit flies from 30 degrees to 18 degrees, you can actually increase their longevity by more than six-fold. I think this is a relatively trivial result, but yet I think it's something that we need to bear in mind when I talk about some of the really spectacular modern advances in increasing longevity, which is, it's this refrigerator effect that may be responsible for part of it, in which case, it's probably got less relevance to human aging than we might otherwise hope.
However, increasing empirical evidence suggests that all of these disparate processes that we had no theoretical reason to assume would be connected, they now appear to be connected. And I'll give you just two reasons why that seems to be the case. First is that simple environmental treatment, such as reducing food intake slows aging in laboratory animals. And when I talk about slowing aging, I'm not simply talking about making them live longer. I think it's easy for us to focus on increased longevity. It's a sort of shorthand, but it often stands, or it should stand for retarded aging. Quite often, we don't know if we've simply increased longevity, or if we've actually slowed aging rate. But in the case of reducing food intake, we have slowed aging in just about every way we can measure, in the rate of memory loss, rate of activity loss, the rate of immune system decline, a whole range of things that we had no reason to think would be synchronized previously.
And finally, the most astonishing, I'd say results, scientific results of the last 25 years, that it turns out that we can alter single genes in a genetically complex organism, and by doing so, increase longevity and preserve functionality to an amazing extent. And I'll talk about what exactly the genes do, and how amazing an extent this is a little bit later, so I think these things taken together, because there's more connection between all of these various processes than we ever had reason to think before.
Now I was asked to sort of describe some of the major branches of aging research, and I've tried to outline them in this diagram here. But let me say right off, it's no longer as easy to do this as it used to be, and that's because the tools of cellular and molecular biology have made all of these disparate fields fuse together. They really now, the techniques that are used for one are used in the other, and so this is more a theoretical construct, than an actual description of various sorts of research that goes on in laboratories.
But starting at the top, the top and going clock-wise, neuroendocrine mechanisms, cellular population dynamics which includes stem cells and telomeres, I'm assuming that this Committee has heard all that they need to hear about those issues, so I'm not going to talk about that.
Then there's organ-based investigations, which are really generally associated with specific diseases, and so I'm really going to focus on the fields that start at 6:00, and go until about 10:00, which is process-based investigations, genetic manipulations and caloric restriction. So there are large fields of aging that I'm not going to mention. I'll be glad to take questions about that, but I think I've got enough to cover just with the other information, so let's start off with caloric restriction.
This is simply reducing food intake. It's not the same as malnutrition. These experiments, usually care is taken to provide plenty of essential nutrients, but simply reduce the level of food intake to about 60 percent of that that animals would eat, if left to their own devices. It's been known since 1935 that it retards aging in many domains in rats and mice, basically, and lots of invertebrates now. That is, it really does affect activity, neurological decline, immune decline, just about everything that we can measure. It's by far the best described of the age retarding treatments because it's been studied for so long.
An important point is that it's not due simply to reduced metabolism. The animals that eat less become smaller, and if you calculate their metabolic rate per cell in their body, this is not the refrigerator effect. They're actually processing energy at just the same rate as the animals that are fully fed.
One of the striking things that you don't often read about these calorically restricted animals, is they have enormously increased spontaneous levels of activity; whereas, a young mouse might run a kilometer in a night, a caloric restricted mouse might run six or seven kilometers in a night. Whereas, a normal mouse would stop running at all by the age of eight months, these animals are still running several kilometers a night at the age of two years, so they're enormously different in terms of their activity. However, it's been well-documented that the retarded aging that we see in caloric restricted animals is not simply the affect of exercise. Exercise does something, but it doesn't do this. It does something different.
The other thing is that it's been very difficult to investigate because it changes so many things physiologically. Literally hundreds of changes in the body occur, at least in mice and rats, when you simply retard their food intake. However, the new technologies that have arisen in the last decade or so promise to facilitate our future understanding of this. And I think now that the prospect is in view, that we will really understand how this simple environmental treatment slows aging as dramatically as it does.
Of the process-based areas of research, I'm going to talk about one that has the most press, which is oxidative damage, and also probably has the most unanimity of agreement among scientists about its central importance, so just a couple of words of background.
There is indirect evidence from all sorts of sources that supports the view that oxygen-free radicals which damage basically every biological molecule, and which are produced as inevitable consequence of eating and breathing, cause gradual deterioration of lots of cells and tissue. Our bodies have all sorts of anti-oxidants that are produced by our cells that destroy many, but not all of these oxygen radicals, and is probably a slight imbalance between radical production and anti-oxidant activity that may modulate the impact of longevity enhancing treatment; such as, caloric restriction. It's not always that there is an increase in the activity of anti-oxidants, but we generally either find an increase in the activity of anti-oxidants, or a decrease in the production of oxygen radicals.
So now some research findings that I think bear on the importance of this issue. Of all of the genetic mutations that enhance longevity in worms, these small nematodes that are one of the models that biomedical researchers now rely on to investigate natural processes, one specific anti-oxidant seems to be critical for the effect of all of these. If you get rid of that particular anti-oxidant, you get rid of most of the effects.
There's also a study published a few years ago which showed that a synthetic anti-oxidant made by a private company, that was to combine the activities of several naturally occurring anti-oxidants, extended the life of these worms. If we genetically enhance anti-oxidant activity in fruit flies, it's well documented now that that extends their life. And we know that caloric restriction in mammals reduces the production of oxygen radicals.
And finally, long-lived mouse mutants, these are genetically altered mice that live up to 50 percent longer than standard mice, have enhanced anti-oxidant activity, so all of that suggests that oxidant production, oxygen radical production, anti-oxidant activity are keys to understanding this. However, despite the success with invertebrates, when we make mice that were genetically engineered the same way that the fruit flies have been engineered to live longer, they don't live longer. It hasn't worked, so there hasn't been a translation from the worm and the fruit fly biology to the mouse biology at least in this one domain.
But on the other hand, I'm going to describe an unpublished experiment that I have permission from the experimenters to talk about. Recently, where they've really done something that's not at all natural. They've manipulated anti-oxidant in an extremely novel way, and really have produced a mouse that lives longer because of enhanced anti-oxidant defenses.
Now the part of aging biology that has the most excitement currently is aging biology which is retarding aging by the identification of single gene mutations, very simple changes in the genetics of complex organisms that lead to enhanced longevity, and what has been looked at, increased function.
The significance of identifying these single gene alterations is that in doing so, it allows us to relatively precisely trace the biochemical pathways that are responsible for the change in the aging rate. The huge surprise -- I used to say that we'll never find a single gene that changes aging rate in any organism that's composed of more than 1,000 cells. And the chief model in this, the worm is composed of 959 cells. That was why I said that. It turns out I was dreadfully mistaken about this. In those worms we now know of mutations in more than 50 single genes that lead to increased longevity, and they're just about thoroughly done with scavenging the genome, the worm genome which has about 19,000 genes. And it looks like ultimately there's going to be about 200 genes, changes of which in any one of them lead to increased longevity, so it's a dramatic number from my perspective.
It also turns out that we now know of six genes in mice that do the same thing; that is, an alteration in any one of these six genes extends life, and in mice we know preserves function quite dramatically. Now that sort of investigation has only just begun in mice. And I'm sure as that 50 to 200 genes are basically identified in mice, and are altered, that number will grow, as well. So an enormous number of simple genetic changes have led to increased longevity. The other thing is that the effects on longevity have been astonishingly large, and in a second I'll tell you exactly how large they have been.
And the final thing is that it would be easy to make the case that these are worms, these are fruit flies, these have been evolutionarily diverged from humans for more than a billion years, so what possible reason is there to assume that these changes have any relevance to humans? And that's clearly an open question, but there's at least some suggestive evidence that they might be highly relevant to humans. I'm going to give you a little Alphabet Soup here, and I want you to ignore it, but this one pathway which makes insulin in humans and makes our body respond to insulin, and also another molecule which is an insulin-like growth factor. In that particular biochemical pathway, there's enormous preservation of that pathway between worms and fruit flies and mammals, and what those letters mean is unimportant. What the important point is, that if you'll notice the asterisks there, those are points in this biochemical cascade where alterations lead to substantial increases in life, and so have alterations in similar genes, in organisms that have been diverged for a billion years or more, that all extend life. And that suggests that there, indeed, may be highly conserved general mechanisms, and that this may ultimately have relevance for humans.
So let's focus on the worms where the dramatic developments have really taken place. So how extraordinary has been longevity extension in worms? Worms have 100 million nucleotides in their genome approximately. If you change one of those, one of the letters, one of the 100 million letters of the genetic alphabet, you get a doubling of life span. If you change two of the letters, if you change one letter in one gene, one letter in another gene, you'll get a tripling of life span. We so far have nothing of this magnitude in mammals, not caloric restriction, not genetic mutations, nothing.
Another key feature is that these mutations are effective at increasing life span if they're only activated in the nerve cells. Now one of the reasons we work on these worms is because they're so well described. And when I say they're well described, here's how well. They have 959 cells in the adult animal that are not eggs or sperm, 302 of those are nerve cells, 131 cells have died during the course of development. I mean, we know this animal really, really well. And so if you only have these mutations affect the nerve cells, you get the same life extending effects.
The other thing is there are some differences. Basically, we don't see changes in the nerve cells, unlike ourselves where some of the most dramatic changes are nerve cells, the worms usually seem to die of changes in muscle pathology, so that suggests that there may not be certainly a one-to-one correlation between what happens there. There's no obvious nerve aging.
Now I think to place this in a context, you need to understand a little bit about worm biology, because it's very different than mammal biology. These worms are about the size of a comma in your briefing booklet there, so they're very tiny. They go through four larval stages, and then become adults and live a couple of -- you can see the pictures of young worm and an old worm there. They tend to get superficial wrinkling as humans do, as well.
Now a key factor is that little diagram on the right there, which shows that they really go through this sort of time-out phase, a phase that you could think of as a hibernation phase. They don't become inactive, but they stop feeding, they start moving around a lot, and they basically stop aging for up to months. And they do that when basically they run out of food, and that's a very, very different sort of biology than humans have.
The other thing is when they're adults, they basically are having no cells that continue reproducing. And sharks do get cancer, unlike the book of the popular title, but worms don't get cancer. And that, again, makes their biology dramatically different than the biology of mammals. So one of the genes that's been most studied is this gene called daf-2, which is an insulin receptor. We have a very similar gene in humans. In fact, this gene is about 50 percent conserved in its sequence of amino acids that make it up with the human one.
A single nucleotide change doubles longevity, as you can see. Now most of the worms, just another aspect of their biology, most of them are hermaphrodites. They don't have males and females. They have male and female in the same organism, except about one out of 500 is a male. If we look at that same genetic mutation in males, we get more than a six-fold increase in longevity. Something on a scale of increase that I think nobody was prepared for.
What's more, if we compare some of the simpler mutations with feeding them less, we also get about a six-fold increase. In all of the experimental animals so far, we don't seem to be able to push much beyond a six-fold increase in longevity, but we've achieved a six-fold increase by several means. And mutations in this same gene in fruit flies and laboratory mice also lead to extended longevity, although not nearly so dramatically.
In mammals, and I think here we feel safer extrapolating from one mammal species, that is mice, to another, we know that among the four -- the six genetic mutations are four that are all involved in this insulin pathway. If we take those mutants which are dwarfed mice, and we furthermore calorically restrict them, we increase their longevity by about 75 percent. This is the best we can do by mammals, but if you think about increasing human longevity by 75 percent, I think you realize we're talking about something really dramatic.
Now let me make a few points about these genes, because they get reported in the press as being uniformly good. Isn't this wonderful, but there are really some things you should know. Virtually all these genes reduce normal gene function; that is, nature has produced a gene that presumably has a utility and how we mainly extend life is by disabling those genes.
A consequence of that is that all of these genes really do have substantial side effects. They don't only increase vigor and increase longevity. Some of the common ones are either sterility or reduced fertility, and that's also true of the calorically restricted animals. Some actually decrease longevity in one sex, although it may increase it in the other sex. This is particularly true of some of the genes that have been reported to extend life in fruit flies. They almost all reduce body size. They increase susceptibility to cold, which is relevant in the real world presumably. They also reduce competitive ability; that is, if you take animals of the increased longevity-type and you put them in the same arena with normal animals, and you allow them to just reproduce as they will, pretty soon the longevity genes will disappear from that arena. And none of these genes has ever been identified in natural populations. These are not things that we get out of nature. These are things we create in the laboratory. They do well in the laboratory. They would probably not do well in nature.
Let me just give one unpublished example of a gene that I think is the sort of thing that we're going to see more and more of. This is one of the few genes, like I say, they're just a handful and increased longevity comes from an enhancement of a normal gene product, so this is a mouse that's been genetically engineered to over-produce a cellular anti-oxidant. Anti-oxidant called catalase. What makes it unique is that it's been directed. It usually is active in one part of the cell, it's been redirected to another part of the cell, this part of the cell that produces most of the oxygen radicals. And when you do that, we now have the first mammal ever where increased anti-oxidant activity increases longevity. There haven't been a lot of functional studies. We don't know if it's just longevity, or if it's a generalized retardation of aging, but we do know from looking at the animals that died, that the sorts of normal mouse heart pathology has decreased in these animals.
Now very quickly, the last two questions, will average American life span significantly increase in the future? Some of you may know that Jay Olshansky and I have a wager of about a half a billion dollars on this issue, so I'll present my biased impact. First of all, being a scientist, I never say anything that I can't take back, so I'll say it depends on what you mean. You can see I've been influenced by political events here. It depends on what you mean by significant.
I think in the near term using traditional medicine, I think Dr. Olshansky and I would agree that we will get a few additional years, whether it's going to be five, whether it's going to be ten, whether it's going to be three, I think we'll find out. Longer term as the sorts of therapies that I've been talking about that work in animals actually get extended to humans, and I think that some of them will end up being relevant. Many of them will not be, I think it's easily possible that we'll get a few additional decades of human life expectancy.
The second part is, depends on what you mean by a year. And let me just give you my best guess about this. I think the time horizon for anti-aging therapies is probably something on the order of 30 to 60 years from now; that is, anti-aging therapies that really do work, and really do work in humans.
The last question, are there hard physiological barriers on the maximum human life span? I will say we don't know, but I will show some animal data that I think suggests that there are not. Let's imagine this is a hypothetical animal survival curve, and this is some sort of extrinsic physiological barrier to life span. One possible way to ask this question is to say, what if we could increase animal longevity and bring it right up against that life span? You'd basically have a more square curve, because more animals would be dying closer and closer, and closer to whatever that intrinsic limit is.
Now over the past few years, there have been these caloric restriction experiments which, like I say, universally increase longevity, and we have actually a chance to compare curves that look like these two curves. And here are curves that are just from four of these studies, and if you look at the one in the upper left, it does sort of look like there might be a barrier; that is, the slope of that curve towards the end is really more steeply defined for the restricted animals. But if you look at some of the others, you'll see either no such trend, or you'll see that it even looks like there's a less steep slope, so I would say on the evidence of what we know about animals, we don't see hard fast barriers. And I would also suggest if we can change a single genetic letter out of 100 million and double the life span of an organism, it suggests there aren't hard barriers, as well. And that's all I have to say, and thanks for your attention, and love to discuss what I had to say.
CHAIRMAN KASS: Thank you very much. Could we get the lights so we can actually see each other. Thank you for a wonderfully clear and synoptic presentation. The floor is open for discussion. Dan Foster.
DR. FOSTER: Just one quick technical question. In the unpublished experiment on the catalase activation and the expression other than it's normal expression, was that to move it into mitochondria for the electron transport?
DR. AUSTAD: Yes.
CHAIRMAN KASS: Rebecca was it, Rebecca Dresser.
PROF. DRESSER: I was wondering if you could say a little bit more about how these animals die, animals that have had their longevity extended? At the very end, you were getting to that in your presentation and the readings. Most people say the idea is well, we don't want to just extend the life span, but we want to extend the high quality of life prior to death, so has anybody looked at -- I guess the closest thing would be the mice, how they die? And I guess another concern would be something like dementia or other subtle effects of human aging that are of concern, and are things that most people dread, would be difficult to detect in animals. If we were -- if our goal were to extend the human life span and make it of a high quality, and avoid the catastrophe you mentioned of dementia, how much of that could we learn from animals?
DR. AUSTAD: Well, okay. That is a very good question, so let's just focus on the calorically restricted rodents, because we know so much about those. Caloric restriction does two things. It gets rid of some diseases entirely, and it postpones others. From what we can tell -- the other thing that pathologist, after pathologist, after pathologist tells me who's looked, who's done autopsies on these animals, about a third of them you find no pathology at all, so it's not clear why they died. It's not that they have lots of small things wrong with organisms, at least from what you can see at the level of the microscope. In a fair fraction of them, you see nothing whatsoever.
Given the fact that they maintain their activity to so late in life, it suggests that they're living really a highly functional life. Now in terms of dementia and whether we'd be able to see anything like that, there actually are ways of investigating this in mice. You know, and I think it's a pretty good investigation, because one of the things with Alzheimer's Disease, for instance, that's clear, is that there's a dramatic decline in spatial learning and memory, and that's very easy to kind of work out in mice. And when you do that, you do find a reduction in the rate of decline in spatial learning and memory, as well. So certainly it's not a perfect analog of what we would like to see in humans, but it does suggest that you really are pushing back neurological problems, as well.
CHAIRMAN KASS: Elizabeth Blackburn.
PROF. BLACKBURN: A related kind of question too, and perhaps either speaker could help me on this, in terms of the same question. But what do we know about it in terms of human clinical information? In other words, if we take those people who do, you know, quite naturally live longer lives versus those who have lived shorter lives, and look at the duration and intensity of their age-related diseases. Do we see that there's simply a postponement then producing the same spectrum and duration of age related diseases, in those individuals who have lived for a very long time, as opposed to those who have lived for a shorter time? Let's say, you know, 80 year old versus centenarians. Are there data about that question, which I think would answer Rebecca's question in part too.
DR. AUSTAD: Well, this is something that Jay probably has at his fingertips better than I do, but let me -- my impression is that certainly people who live exceptionally long lives, let's say centenarians, really do seem to be protected from some diseases; for instance, cancer rates are much lower. Some things continue to increase as, you know, as old as people get, but some things do tend to just not occur in those people. Jay, do you want to --
DR. OLSHANSKY: Yeah. You'll see the same pathology in the extreme elderly than you will in the elderly. Everything is postponed. There's something very unusual about the centenarians and the super-centenarians. Somehow, whatever it is genetically that enables them to combat heart disease and cancer for 100 or 110 years is what makes them so unique. But in terms of pathology, you'll see pretty much the same things in the extreme elderly.
PROF. BLACKBURN: I was asking this because I think this is going to relate to the bigger question of, by postponing when life ends, is that going to change the medical burden, as one would say, to society, or simply move everything into the later stage? And I just didn't know if there's information that would suggest one way or the other.
DR. AUSTAD: Yeah. That's actually been a highly controversial issue among demographers, and I'll defer to Jay to sort of answer this question.
DR. OLSHANSKY: Yeah. This issue about whether or not we are altering the expression of frailty and disability among the older population as we extend life is one that researchers have been attempting to get a handle on now for the past couple of decades. And I think initially, the belief was as we were extending life, we were making -- we were increasing frailty and disability among the older population. Some of the more recent data by some researchers at Duke and elsewhere, suggests that we may be improving the health of the older population as a result of the extension of life. And I wasn't -- I mean, I had a really nice figure. I wasn't going to present it, to illustrate that great caution in our effort to extend duration of life because of the very fear that Steven talked about at the beginning, if we push people out into an age window in the life span where the expression of frailty and disability is extraordinarily high, then we will face a scenario that we perhaps don't want to see, a trading off of longer life for worsening health, which is the motivation for research on aging, to postpone the frailty and disability into later ages.
DR. AUSTAD: If I could just add to that, I actually had an e-mail from some of these people at Duke last night with some of the latest information, saying you might want to present this tomorrow. So I didn't put it in my presentation, but I'll just say it's a continuation of this idea that we're actually decreasing frailty at specific ages, you know, fairly dramatically right now. And it's happening on a very small time scale; that is, the rate of decrease in frailty seems to have accelerated even since the early 1990s, so there's more stuff that's going to be coming out.
CHAIRMAN KASS: And this doesn't have to do with changing health habits and exercise?
DR. AUSTAD: Oh, it very probably does have to do with that. It probably has to do with increased medical surveillance and better treatment. Yeah, absolutely.
CHAIRMAN KASS: Janet Rowley.
PROF. BLACKBURN: Can I take advantage of my working mic to sneak in a question?
CHAIRMAN KASS: Janet, is it all right, we'll let Elizabeth go? Please.
PROF. BLACKBURN: Well, it was more a comment to enlarge on, I think your very appropriately cautionary note about saying that when you mess up longevity genes by dropping their function down a bit, and increasing longevity, one sees that there are costs to this. But I think it's interesting that some of the recent information from the worm has shown that there are clearly mutations which do increase longevity in those dramatic ways you described. And there is no measurable affect on sterility or ability to produce healthy progeny at all. So in a few cases, there is an uncoupling. Now I realize that there are many cases when there's not, but the point is that one can start to see, at least, a case where there is not an apparent cost by what's being thought to be the traditional linkage that sterility and fertility are problems that sometimes are the converse side of extension.
DR. AUSTAD: Sure. If I can just comment on that, I -- that's only one of the side effects, the sterility/fertility.
PROF. BLACKBURN: Right.
DR. AUSTAD: There was another mutation that for years was known to be the cost-free mutation. And we found the cost of that as soon as we put the worms together with the normal worms, and found that they weren't -- and that hasn't been done with this particular one, so we haven't found any cases where there's a cost-free addition to life yet, just because of inadequacy of the investigations to this point.
CHAIRMAN KASS: We have to take a pause for five seconds to get the system working, a moment of meditation. Janet, please. Janet Rowley.
DR. ROWLEY: I have several questions; one of which is fairly straightforward. I assume that the six genes that have been identified in the mouse are homologous to some of the genes that were found in the C. elegans?
DR. AUSTAD: Yes. Four of the genes are all part of the same pathway that was the original one that best describes C. elegans. The other two are different genes. They certainly have homologs in C.elegans, but I don't know what their effect is yet in C.elegans.
DR. ROWLEY: Well, the thing that astonished -- well, the two other things that I think merit some comment, you said that the mice that were calorically restricted exercised more. Now I would have thought that that would actually increase the amount of free radicals, at least in muscle, so you think that that would be a deleterious effect. I also understand, of course, that exercise has its benefits, but overall I gather that the free radicals don't increase in calorically restricted mice.
DR. AUSTAD: No. Calorically restricted mice have been looked at in some detail in this, and originally people thought that there would be a sort of increase in all the anti-oxidants. That's turned up to be highly variable, some tissues yes, some tissues no, some anti-oxidants yes, some no. But what's universally the case is that if you look at free radical production, it's decreased in these, even though they're exercising a lot.
DR. ROWLEY: Okay. And the last question that I have relates to your statement, and I may have misunderstood it, about the mutations and you needed mutations only in nerve cells to get this increased longevity, so that all the other cells in the body either don't have this mutation, or it's not active in those cells. So what's so special about having it in nerve cells? It seems to have a generalized effect.
DR. AUSTAD: Well, this is in the worm now, C.elegans.
DR. ROWLEY: Right.
DR. AUSTAD: It's probably basically because it improves coordination of all processes throughout the bodies through hormones. I mean, the guess is that this is a so-called neuro-endocrine effect, because they've done it exactly that way, if you only allow these genes to be turned on in the nerve cells. And this was not anything that an a priori prediction. It was just an empirical and somewhat surprising finding, I would say.
DR. ROWLEY: And is it the Sir-2 gene or related to it, or is it different?
DR. AUSTAD: These were daf-2 genes.
DR. ROWLEY: Daf-2.
CHAIRMAN KASS: Michael Sandel.
PROF. SANDEL: I want to make sure I understand something you said early in your talk. In the briefing papers that frame the part of the discussion later on the ethical and social implications of this, a contrast is drawn between enabling more people to reach the natural limits of life, and pushing back those limits; that is, increasing the maximum human life span. As I understood you, over the course of human history, both of these things have dramatically changed. Is that correct?
DR. AUSTAD: Yes. Both of them have dramatically changed, but the implication is not that there's been a change in the biology of humans, but simply latent possibility for longer lives has been revealed by basically a friendlier environment.
CHAIRMAN KASS: Frank Fukuyama.
PROF. FUKUYAMA: You haven't said anything about the body of theory that comes out of evolutionary biology on aging, and I understand there -- I mean, that's less well-established, and there are a lot of controversies about, you know, why you have post menopausal women and so forth, but out of that body of theory, is there, you know, a consensus, for example, about why human life spans, you know, are what they are, compared to other species, what adaptive significance that had and so forth?
DR. AUSTAD: Oh, I'm so delighted you asked. This is really my specialty, and that body of theory is actually extremely well-supported by a huge mass of empirical evidence. And the basic idea is that the less one is subject to extrinsic hazards, to environmental dangers, the greater will be the selection for maintenance of the body. And this makes a whole variety of predictions, not the least of which is that when you create a safer environment, as we have done for ourselves over the last thousand years or so, that ultimately, you're going to get the evolution of a longer life span. So even without any medical advances, all the empirical evidence suggests that over the next 25 or 30 generations, humans will biologically live longer. Our limits where they are will be increased, because of this evolutionary theory, that's its prediction, but it's also been validated by exactly those sorts of experiments in the laboratory, really dozens of times in fruit flies.
And also another thing that we predict is that there may be things that age more successfully than humans, and that's clearly the case. There are things that manage their oxygen radicals much better than humans do. And one branch of research that I didn't talk about, because it happens to be my own branch of research, has to do with understanding how those animals that do it better, do it.
PROF. FUKUYAMA: Just as a follow-up question, I mean, is there a theory about why humans live as long as they do compared to other species?
DR. AUSTAD: Yeah, the idea is that basically a combination of their social environment and their intelligence has allowed them to avoid a lot of environmental dangers that other animals -- primates as a whole are a long-lived mammal, and they're a long-lived mammal it's thought because they're exceptionally intelligent, and they're exceptionally social. And those two things together will allow them to avoid a whole host of environmental dangers that let's say small things that creep around in the dark would not be able to avoid. And the best evidence for this sort of reduction of extrinsic mortality probably comes from the fact that animals that fly, whether they're birds, or whether they're bats, are exceptionally long-lived for their body size in all examples. For instance, a little brown bat that's a fraction of the size of the mouse has been reported to live over 34 years in the wild, something that is absolutely unheard of in non-flying mammals.
CHAIRMAN KASS: Gil Meilaender.
PROF. MEILAENDER: I don't know if this -- I mean, my question really I think is whether what I'm about to ask makes sense. You could picture at some point farther down the road two kinds of possibilities. One, that nobody ever died from some sort of sudden pathological problem, but just wore out eventually, or alternatively, you could picture a future in which nobody ever really wore out, but there were still sort of medical crises that caused people to die.
Is it conceive -- can we actually separate those two. Are the two related in such a way that it makes sense to think about those as possible futures or, in fact, however we extend longevity, do some people die? Will some people die by wearing out, and some people die because of some -- even though they're not worn out, because of some sudden crisis?
DR. AUSTAD: The way I interpret your question, I think you're basically saying will we be able to slow aging, or will we will able to stop aging? And there's certainly no work in any animals that suggest that we're anywhere close to stopping aging in any organism at this point.
We've learned how to slow it dramatically, and what we're doing there is we're basically the animals are still wearing out. They're just wearing out at a slower rate.
PROF. MEILAENDER: So you wouldn't picture a future in which, however we died, we all died sort of in the pink of health.
DR. AUSTAD: That would be the ideal, certainly. And this is one of the things that the demographers have been arguing about, as we are living longer and longer, are we really compressing the period of ill-health or not, or are we simply just moving it back? And like I say, that's a controversial issue, and I bet that Jay was going to cover that in his talk anyway.
CHAIRMAN KASS: Paul McHugh, and then I'll put myself in the queue.
DR. McHUGH: I found it fascinating what you were saying, but all the things that you talked about, everything you showed was steadily declining as we grew older. And you showed that in animals, as well as in human beings. But, you know, some things get better with human beings as they get older. They get smarter. They don't run 5,000 meters any more. They walk. And they write better poetry probably, they do a number of things better when they're old than when they're young.
What is the development in your work and in your study of human beings that would begin to emphasize these healthy sides of aging, or these good sides of aging? Perhaps this is not appropriate to ask you, but this comes up a lot in psychiatry, particularly a psychiatry of meaning is always ready towards any discouragement in old people as, you know, this is what happens when you get old. I remember these discussions quite vividly when I first began in geriatric psychiatry, and I would say, you know, this person is depressed. She needs some help, and I would get back wise remarks about well, don't you understand what it's like to be old, Paul? You will understand it some day. You lose everything. You grieve, you give up your functions. And I said, well, why don't we give this person a little medicine, and then they became more like old people who are happy, and who are engaged in things, who are politically alert, and have wisdom.
Where in this research is going to come these things which fundamentally for me, anyway, if there's going to be a discussion about longevity, it's these things that are wanting to be preserved, not necessarily for me, but for the people I love, and for the people I cherish, and people whose gifts to my life I would like to preserve?
DR. AUSTAD: Yes. There is a concept called successful aging that makes biologists very uneasy, because like I say, there are dozens and dozens of things that all decline. And when I talk about aging, people often say well, that's so discouraging. Does everything decline? And I say no, everything does not decline. There are certain things like, you know, density of facial wrinkles, that goes up, but actually, one of the things that does increase, and you mentioned it, is wisdom. Sort of raw measures of mental function clearly decline, but there are things that if appropriately measured, like wisdom, the ability to see multiple solutions simultaneously that increase. And I would agree that that's incredibly important for the human condition, but it's something as basically an experimental biologist, that's very hard to assess in animals, and so I usually don't talk about it, but that's very clearly the case. And one of the things about preserving mental function is that that would be an important aspect of it.
Certainly, when I worked in New Guinea among, you know, very remote areas, the elderly people in those areas are treasured as a source of information. In fact, I went in to some areas that no one had been into for decades, only because the elderly people in the village still knew how to get there, so you're absolutely right. It's important. It's hard to talk about though in animals, which is my specialty.
PROF. SANDEL: It's probably hard to know whether those fruit flies in the refrigerator are growing wisdom.
DR. AUSTAD: Yeah, exactly.
CHAIRMAN KASS: A comment, and then a question. First of all, I would at least like to register as a question whether our immediate intuition that it would be absolutely best to go in the pink of health is, in fact, correct, and whether or not -- we talked about this a bit last time, and whether death would become simply intolerable if all of us were whisked away without any kind of anticipatory decline. And one certainly couldn't comfort the survivors to say that someone was released from their decline. That's the philosophical and ethical question for reflection and discussion.
I take it that unlike Dr. Olshansky, if I understand his writings correctly, you seem to be suggesting that there is some -- that there really is some kind of genetically determined species-specific life span, at least that this isn't just somehow the accumulation of post reproductive errors, but that the genetic results seem to suggest that there's some kind of switches that control something like what is the built-in possibility.
Second, you seem to say -- you said that we don't find any of these mutations in nature, which would suggest that they would be disadvantageous. If these are indeed somehow connected to longevity, they're disadvantageous. An the third thing you said was there seems to be some retardation of reproductive prowess and decreased competitiveness. It's my sloppy language, you were more precise.
This question has to do with the connection between longevity and fertility, and the way in which -- what the implications might be if these things were really deeply linked, that thinking about altering the genes that would produce greater longevity would not just biologically, but perhaps even also culturally, separate question, decouple these things, or at least produce a kind of premium on long-lived, non-childbearing creatures. So forget the editorializing and deal with the biology. Is there some kind of -- how do you explain the fact that some of these things don't crop up? You could say that there would be no selective advantage, perhaps, but there seems to be a selective disadvantage for the appearance of these things. And does that have something to do with the fact that to make the world safe for progeny, the old ones have to die?
DR. AUSTAD: To this last point, I would say yes and no. I think clearly these things are -- would be disadvantageous in the state of nature, either because of their affect on fertility or their affect on developmental rate, or their affect on competitive ability. For instance, we know in the worm that it's the ability to come out of this dauer phase that's slowed down, so I think for every one of these would be disadvantageous in nature, and that's why we don't see it. The question is, do we live in a state of nature in which something that changed, that for instance, might influence our fertility substantially. Let's just imagine that we knew how to slow it, we had a pill, that if we took the pill that would make our fertility much, much lower than it is. We could simply choose to take that pill later in life, presumably. In the calorically restricted animals you still get an effect if you start the caloric restriction later in life. Now passed a certain point, you don't, and you get a smaller effect. But this impact on reproduction seems to be important, but not critical. There do seem to be instances where you can tease them apart. For instance, if you look at a mouse that's a year old, in which case they're pretty much post reproductive, you start it on caloric restriction, you still get a small effect.
The philosophical question about whether it's best to go in the pink of health is a very interesting one. I think there's a quote from Montaigne I think in your book about that. That is something I hadn't ever thought of before, but I thought was a point worth pondering.
CHAIRMAN KASS: Thanks. Bill Hurlbut.
DR. HURLBUT: Just a tiny footnote. I mean, in addition to Paul McHugh's comment about the question of wisdom, some of the benefits of the aging process, it also seems to me that the question of the link of this to reproduction means that to take simply a medical and health-related view to this, and not to see its implications for the relation amongst the generations is also, I think, a limitation on the perspective. I'm not accusing anybody here of taking that, but I think it has to be added. Bill Hurlbut.
DR. HURLBUT: I want to ask you a little bit about some of the implicit assumptions that are going into your projection here. When I think about the models that you're studying like C.elegans or in some other cases fruit flies, I'm very struck right away by how human beings are not the equivalent of just big fruit flies or long living roundworms. And I wonder specifically in those models, for one, they may have been selected for study because they have tight genetic controls, and therefore, are more determined organisms, perhaps. And also, in the case of roundworms, the adult form, at least is a post-mitotic existence, no more cell divisions which changes the scene a lot. And so what I'm wondering is, you spoke of the conserved genes that connect human beings to fruit flies as being, what I think your statement was 50 percent conserved, which means there's a lot that's not conserved.
And I'm wondering given the -- notwithstanding some of the evolutionary arguments for why species have this reproductive growth, longevity ratio and so forth, I can see reasons to suggest that human beings are, as a species and as individuals in that species have benefits to longevity, such as the wisdom you spoke of, and there's some evidence in macaque monkeys that the troops with longer-lived members survived better, so is it possible that we've actually already been selected for these mutations, if you will, of longevity. And that, in fact, human beings already represent the genetic form that will allow their maximal or some degree of maximal longevity? You see what I mean?
DR. AUSTAD: Yeah, I do, and I think they're good points. And certainly, humans are not big worms or big fruit flies. And I think we deserve to be skeptical about the translation of those results. And my guess is that most of those results, those 50 plus genes that we already know of, are going to have no relevance to the mammalian case. However, we do know these genes that do similar things in mice, and it strikes me that that offers more hope of extrapolation.
In terms of whether humans have kind of already maxed out what you can do as a mammal, that's an interesting case. And it's basically the issue under whether we can learn anything about human aging from studying animals that by definition are very unsuccessful at aging. Here we are. You know, we live decades in states of high fitness, so what can we learn from something that lives a few weeks? And I think that's again an important caveat, and it's one of the reasons that there's now movement afoot to look at animals that really are more successful at managing some of the damaging processes. And let me just give you an example from my own work.
I work on birds, and I'm looking at basically anti-oxidant capacities of birds. And the reason I'm interested in them, if you calculate the amount of oxygen they process per cell in a lifetime, birds process anywhere from three to five times more oxygen per cell per lifetime than humans do, so they do something better than we do, so it's an alternative approach. I think it's appropriate to be cautious about interpreting extending life in very short-lived things to extending life in humans. I perfectly agree.
CHAIRMAN KASS: Questions, comments? Alfonso Gomez-Lobo.
DR. GÓMEZ-LOBO: I have an information question. Are there species that have a longer longevity than human beings? I mean, how many, and what would they be? I mean, where do we stand in the scale of things? Are we at the upper end, or we in the middle.
DR. AUSTAD: Well, we're towards the top. It depends on how widely you cast your net. If you cast it into the plant kingdom, then we're not up near the top. But if you focus on animals, we are -- certainly one animal that's recently come to life that's much longer-lived is bowhead whales, of all things, where recent evidence suggests that they live more than 200 years. That evidence is indirect. It's biochemical analyses, but it's supported by the fact that bowhead whales that have died recently have been found with traditional hunting implements that have not been in use among the Inuit for more than a century, so suggests that they really do live, you know, upwards of several centuries.
CHAIRMAN KASS: Tortoises.
DR. AUSTAD: Tortoises are also longer-lived, yeah. I mean, there are these things. There are mollusks that have been documented over 200 years. There's a rough-eyed rockfish, don't eat these, that's been documented 205 years, so there are other animals. Ourselves and the whales are exceptional because we're warm blooded, we have this high metabolic rate that we don't have the option of turning off when we feel like it, so that's why I focused on those.
CHAIRMAN KASS: Janet Rowley.
DR. ROWLEY: I'd like to come back to your comments about the fact that because of the complexity and going back to the caloric restriction model, though I wouldn't limit your answer to just that, that we now seem to be on the verge of either having new insights or new techniques, which in the future is going to sort out some of the reasons that caloric restriction leads to longevity, and therefore by implication may be applicable to other kinds of systems. So would you -- firstly, what are the new things that you see on the horizon in your own field? And would you speculate on how some of the new discoveries might be applied in the future?
DR. AUSTAD: Sure. Let me just give you one example of how we can start to investigate this. There have been dozens and dozen of hypotheses about why caloric restrictions works. It's been virtually impossible to disentangle those, but now there are ways to do it. For instance, one of the hypotheses has had to do with the stress response, and the fact that stress is mildly elevated in these animals, and that moderate levels of stress might be good for you. You know, that's a controversial hypothesis, but yet it's been a hypothesis that's been around a while.
There are now ways to genetically engineer mice so that they cannot have that moderate elevation of stress, and so you can genetically engineer them, and you can see does that mimic the affects of caloric restriction, and find out either that hypothesis is supported or it's not. And so that's one sort of thing, is that you can start testing one hypothesis at a time.
DR. ROWLEY: And what kind of stress is being particularly implicated, I mean because there are all sorts.
DR. AUSTAD: Right. My guess would be food stress, you know, for anyone who's dieted involuntarily, but it's simply to measure it by looking at elevations in stress hormones. It hasn't been specifically identified as to what the source of the stress is, but it's hard to imagine that it's not food, especially if you deal with these animals. Because when you walk in to feed them, you know, they're doing pull-ups on the cage, you know, waiting for their food, so my guess is that that's --
DR. ROWLEY: That's specific hormones.
DR. AUSTAD: Yeah, that's -- so specific hormones, you know, it's easy to test those sorts of hypotheses. Now the other thing is with the DNA microtechnology, it's going to be easier to sort of find out are there other genetic signatures for enhanced longevity. And we do the same thing when we change a gene, as when we calorically restrict animals, so I think those things are -- you know, it doesn't make the problem trivial, but I think it makes the problem more tractable than it's ever been before.
CHAIRMAN KASS: Bill May.
DR. MAY: I'm not sure you want to respond to this, but as I understood the exchange between you and Leon earlier, it was a reference to Montaigne, and the whole question whether if one improves health towards the end of life, one tends to increase the sting of death, or whether the process of aging and the sense that one is moving into the twilight, and there are increased burdens right there in the body make one readier to accept one's end. I think that's the discussion that was going on. And I must say, not simply reporting personally, but I think there's a sense of readiness to accept death that doesn't relate to the sense that the twilight is here, and burdens have been increased. But also the sense of having completed one's life, and that may or may not relate to increased burdens and woe of one kind or another. But the curious things for us as human beings, what is it that produces a sense of completion? And in the setting of survival as the aim, one would say that the next generation is born and is now, as it were, on its feet, and so one has completed one's task.
But as life extends beyond the 40 and 50 years, on up to the 80 or 90 years, there's a disconnect between the event of the next generation being on its feet, and the actual life that we go on to live. And we are forced to explore different senses of what it means to have completed one's life, than simply having sustained the next generation to the point that it stands on its own feet. And it relates to the whole question whether the sense of end in human life relates simply to survival, or to other dimensions of flourishing and excellence that are different from the aim of the surviving of the fittest.
CHAIRMAN KASS: Now if I might in a way piggyback on that too, and I didn't mean to interfere with your wish to respond. If one is talking about -- I mean, the main interest of this research, as I understand it, is not adding years to life, but adding life to years, and dealing with the infirmities and the debilities that afflict us. Though it looks as if the solution to the second might actually produce increments of the first, and that these two things might very well be linked.
And one of the things that's being changed in the process is one no longer thinks -- one will no longer think of a life having a built-in shape somehow related to the length of a generation unless, of course, the generation now becomes 40 years because of delayed reproduction and the like. And, therefore, it's -- the question of what it really means to have a complete life, when the boundary is moveable, is no longer somehow given by the life cycle, but has to be created. And that's a problem for which the increases of the life expectancy of the past century have already, I think, given us some indication. Where in Western Europe anyhow, initially would be the extreme example, where the birth rate is down to 1.2 children per women for lifetime, that means half the women in Italy are not having any children at all, and therefore, the account of what a complete life for them will be has nothing to do with producing children who reach maturity. It has nothing to do with producing children at all, so there are -- these changes in the life span have profound sort of cultural changes, and the perception of one's own course of life is somehow altered.
I'm not pronouncing good, bad or indifferent, but these are -- there are big changes that have nothing to do with, or that are independent really of the question of dealing with the infirmities that one would certainly welcome being relieved of. Please.
DR. AUSTAD: Yeah, I was -- my original response was these are very interesting philosophical issues, and my role as a scientist gives me no special insight, and I understand that. However, in response to what you said, I was just last week in England talking with a British demographer about fertility patterns in Britain, and it turns out that during times of plague and times of economic depression over the last 1,500 years in Britain, there have been periods of time when half of the women chose not to reproduce because of external events, so I'm not sure it's necessarily just linked to disorientation from this length in life. It seems to be at least arguable, maybe not.
CHAIRMAN KASS: The time has come for a break. Let's take 15 minutes. We'll reconvene and hear from Dr. Olshansky. Thank you very much. It was wonderful.
(Off the record 10:35 - 10:51 a.m.)
SESSION 2: DURATION OF LIFE: IS THERE A BIOLOGICAL WARRANTY PERIOD?
CHAIRMAN KASS: Can we get started, please? Council Members should have at their seats now a map of where we're meeting this evening for dinner. It's just a block away from the hotel, straight north on 7the Street, and information is there. Again to repeat, warm welcome to Professor Jay Olshansky, who's going to give us the second presentation on the subject of aging research and its implications. Please.
DR. OLSHANSKY: Well, first of all, I want to thank you for inviting me. I think you're going to discover that Steve and I really don't disagree on too many things, but we'll disagree on a couple. And I'm also delighted to hear that there are students in the audience. I feel very much at home with the students in the audience, so it's wonderful that you're here, and I will be speaking to you to some extent.
Now I know the title is, "Duration of Life: Is there a Biological Warranty Period?" That's related to some work that my colleagues and I have done recently, but it will all relate to this basic issue that you have raised in this meeting.
Now these are the questions you have asked me to address, and I'm going to address them all. I'm going to address some a little bit more than others, the issue of life expectancy, first and foremost. I'll spend a bit more time on that, the hard physiological barriers. I'll spend some time on that. Demographic implications of life extension, a bit less. Breakthroughs, and then the position statement on human aging that we published during the summer, I will be discussing this.
Now let's start out with definitions first. I think it's really important that our language be correct, we use the proper language when we discuss these issues, so I'll start out with some basic definitions. Life span is defined as the verified age of death of an individual, which ranges, of course, anywhere from moments after a live birth to the world's record for longevity, in this case I'm showing you a picture of Madame Jeanne Calment, who lived for 122 and a half years. So the life span is the duration of life of an individual.
Maximum life span is the longest life span ever recorded for a species, so again Madame Jeanne Calment would be the maximum life span for humans, but it is one individual in a species, longest lived individual, and this number can only increase.
Life expectancy is - and I'll be showing you an image of this in a moment, it's the average number of years of life remaining for individuals at a given age, assuming that age-specific mortality risks from a life table remain unchanged, and what we refer to as period life expectancy is what's most commonly used. A bit more on that in a moment, because you did ask me to tell you briefly how it's calculated, and actually it's pretty important to understanding the prospective changes in life expectancy.
All right. What life expectancy is not, and this is what you will see often reported in various places. The incorrect definition is the average age at death. It is not the average age at death. Indeed, it is a number that is based on death rates observed for a population, and applied to a hypothetical cohort of 100,000 babies. And I'll show you what a life table is. How is it calculated? All right. So I know -- I'm not going to go through this, of course, but I figure you should at least see a life table, and see what it looks like, because this is the basis for the measure of life expectancy.
Now what I did in the next figure -- incidentally, by the way, I finally found another use for my dissertation which was published in 1984, because this figure comes out of my 1984 dissertation. It was sitting next to me when I was drafting this, so I scanned this, and here's the second use. This is a truncated version of the same table, and I just want to illustrate a couple of things here.
First of all, this column here, this M(x) represents death rates or condition of probability of death over here, and this L(x) represents this hypothetical cohort of 100,000 babies. And you basically apply the death rates to the babies. You generate a number of deaths, and then you subtract it from that to get down to the next age group. Basically, the measure of life expectancy itself is based on something known as person years of life. When an individual lives one year, that's one person. For a hypothetical cohort of 100,000, we estimate the number of person years expected to live. You divide it by 100,000, and that's your measure of life expectancy, so this cohort of 100,000 people, babies born in a given year would live this many person years. You divide it by 100,000, and that's how you arrive at this number, so it is not an average age at death. It is a hypothetical number, and the critical assumption to remember about life expectancy is the underlying premise that we assume that the death rate is observed in a given year, the year in which you're measuring life expectancy will not change for the duration of the lives of the babies born in that year. So, for example, for white female babies born in 1978, the presumption is that when they reach the age of five, that will be their observed death rate.
Well, under conditions of declining mortality, as you might imagine, this will under-estimate the observed life expectancy of these individuals. And just as a way to illustrate, we've actually been able to calculate what's known as a cohort life expectancy for the babies born in 1900 in the United States based on how long they actually lived. And when you compare the cohort life expectancy to the period life expectancy, the difference is about two and a half to three years, so it really wasn't that large in terms of -- I mean it's not 20 years. It's relatively small in terms of magnitude, but there is a difference, just so you know.
Finally, aging versus senescence. When you think of -- the word "aging", I know is used most often, but biologists tend to use the concept of senescence to describe what's really happening. Aging, we tend to think of, at least I tend to think of anyway, more as the passage of chronological time, so we all age at exactly the same rate. But we can senesce or grow old biologically at different rates, and this is a classic example of two genetically identical twins, who appear. This one on the left has Alzheimer's Disease, this one does not, as a way to illustrate that even among genetically identical organisms as a result of stochastic random events that occur, stochastic events that occur during the course of life, we can senesce at different rates.
All right. Why do we live as long as we do today? What led to this first longevity revolution? Why is life expectancy now up in the high 70s in humans? Well, this is -- it's actually fairly straightforward to explain this. For the vast majority of human existence, birth rates and death rates were always extremely high, about 50 per thousand population. Right around 1850, death rates began to decline very rapidly. Birth rates then followed until we reach this point today, which is referred to by many as the fourth stage of the epidemiologic transition where birth rates and death rates are extremely low, about eight to ten per thousand. And this is the reason both why we have population aging, and individual aging. It's a transformation that occurred principally within the last 150 years or so.
One of the consequences of that dramatic difference in birth rates and death rates, we all know is global population growth, one of the areas, one of the reasons why I came into this field to begin with, but we also have dramatic population aging. This is referred to as an aging pyramid. It's nothing more than a snapshot picture of the number of people alive at any given age, in a given time period. This is for the entire human population in 1900. It was pyramidal. This is fairly common that you see among most forms of life, a large number of young, and a very few make it out to extreme old ages. And now in the developed world, we have created a much more tectilinear or square like age structure where it is looking -- it's no longer pyramidal. You don't really see this anywhere in animals living in the wild. It's really unique only among human laboratory animals, zoo animals. That's about it.
So why did life expectancy rise so dramatically during the course of the 20th century? Well, if you remember that hypothetical cohort of 100,000 babies that I was talking about before, if you plot out the ages at which they all died based on the death rates observed in a given year, you would get something referred to as a distribution of death. It was the D(x) column of that life table. And if you plot them all out, you get something that looks like this. The area under the curve is the same. This is the distribution of death for U.S. females in 1900. This is the distribution of death for U.S. females in 1985. It is a classic illustration of a redistribution of death from the young to the old.
In other words, the vast majority of this increase in life expectancy from 47 to about 77 through 80 now for females in the United States, is a result of declines in infant, child and maternal mortality. That can only be achieved once for a population. Once it is then achieved, the only way to achieve another increase in life expectancy like that is to influence the elderly. It's a totally different ball game.
I decided to show you this figure to illustrate changes that have occurred in these various parameters. This is just based on U.S. data. Maximum life span based on U.S. data appears to have increased a little bit. I know that there was mention earlier of dramatic increases in maximum life span. I really disagree. I don't think there have been dramatic increases. There have been fairly moderate increases, and I'm not even sure if there have been increases, quite frankly, because we can't really measure the age of everyone on earth. And so it's possible that a thousand years ago we may have had people over 100. I don't know that for certain.
Modal age at death has increased from 73 to 88. Period life expectancy increased from 49 to about 80. This is just for U.S. females, just to give you an illustration of how these numbers have changed as a result of this transformation in the distribution of death.
This figure illustrates the proportion of the gain in life expectancy associated with changes in death rates at different ages. Again, it's a classic illustration, 1900 to 1910. The majority of the gain in life expectancy was associated with declines in mortality at younger ages. And now between 1990 and 2000, the majority of the gain in life expectancy is associated with the population over the age of 45.
All right. Years ago, my colleagues and I published this article estimating the upper limits to human longevity at about 85, and when we published this article, a number of people disagreed with us very openly, saying no, no, no. Things are going to go much, much higher than you anticipate. Our projected life expectancy at 85 was 88 for females and 82 for males. And it will be evident to you why we came to that conclusion, and then we followed that up with a piece ten years later. What we decided to do was to wait ten years to see whether or not the data were moving in the direction that we had predicted, or whether the data were moving in the direction as others had predicted, which was a much more rapid pace than what we had suggested. And we published our findings last year, which I'll go through very briefly in a moment.
Now previous estimates of the upper limits to human life expectancy have all relied on efforts to answer a single question, and that is, how low can death rates decline. We reversed the question. We refer to it as reversed engineering approach, and instead asked how low do death rates have to decline in order to achieve a life expectancy of anywhere from 80 to 120 years. Fundamentally different approach, and this was the conclusion that we came to. This is one of the figures from our original 1990 article.
Now it's life expectancy at birth on the (X) axis, and percentage reduction in mortality on the (Y) axis required to produce these higher life expectancies. So for example, in 1985 female life expectancy at birth was 78.3, 71.2 for males. In order to achieve a life expectancy at birth of 85, you simply go up the axis and go over. You could see it would require about a 50 percent deduction in all causes of death for females, and roughly a 70 to 75 percent reduction in all causes of death for males, just to get life expectancy up to five.
Now to provide some perspective, we calculated hypothetically what would occur with the elimination of various diseases, major fatal diseases in the population, and the vast majority of all humans die from heart disease, cancer and stroke. And we show here that if, indeed, we were to find a cure for cancer, for example, life expectancy at birth would rise by about three and a half years. Life expectancy at birth would also rise by about the same amount, three to three and a half years if systemic heart disease was hypothetically eliminated. And if we eliminated all cardiovascular diseases, Diabetes and all forms of cancer combined, life expectancy at birth in humans would rise up to about 90. So you have to believe that we will be experiencing rather dramatic reductions in mortality in order to yield these very high life expectancies. A life expectancy of 100 required about 85 percent reductions in all causes of death at every age, so when somebody says life expectancy is going to go to 100, this has to happen in order for life expectancy to go to 100. An 85 percent reduction in mortality at every age, which is the reason why we didn't believe it was possible or likely.
We then followed this up with research ten years later. We used data from three of the longest lived populations, the United States, France and Japan. Indeed, as we had anticipated, there was not a dramatic change in the probability of experiencing these much higher life expectancies, even though there were changes in life expectancies during this time period. We did experience eight-tenths of one year increase in life expectancy at birth for females. I have some other figures that illustrate this better actually.
One of the claims that was made by those who were predicting much higher life expectancies was that death rates would decline at 2 percent at every age for every year for the next 100 years. Well, that's a testable hypothesis, of course, so we took a look at the data to determine whether or not during those first ten years following that prediction, whether death rates would decline by that magnitude. And for the United States, we found if you look at the age group from zero to 99, the magnitude of the reduction was only four-tenths of one year, not -- four-tenths of a percent, not 2 percent as had been predicted by those anticipating much higher life expectancy.
Interesting to point out, there were some increases in death rates in some age groups in the United States between 1985 and 1995. There have been some large reductions in death rates in France and Japan. They are not at 2 percent, but they are significant still.
Another way to illustrate this issue of difficulty in raising life expectancy, life expectancy on the X axis, percentage reduction in mortality required to raise life expectancy at birth by one year. So, for example, when life expectancy at birth is 50, it takes about a 4 percent reduction in death rates at every age to raise it to 51. When life expectancy gets up to 80, it takes about a 9 and a half percent reduction to produce the same one year increase in life expectancy. In other words, and this was really the principal message that we were trying to get across with our original article, was the higher life expectancy goes, the less sensitive it becomes to changes in death rates. The higher it goes, the less sensitive it becomes, the more difficult it becomes to raise the measure further, which is actually the reason why we were suggesting that life expectancy is, perhaps, not a very good metric, a very good measure to tell us anything certainly about the health of the population.
This will be the last figure I'll show on this, because I realize I'm beating you to death with this issue of life expectancy, but this is actually an important one, because one of the questions that came up often after this issue first arose was, well, if you were around in 1900 and you were asked the same question, how high would life expectancy rise, would you have predicted that it would go up to 80? And my answer was no. Well, then I've been told how could you make a prediction today, based on what we know today?
Now actually, as it turns out, believe it or not in a way this is a testable hypothesis. The X axis is age, on the Y axis is the proportion surviving. Remember Steve talked about this earlier. This is the same hypothetical cohort of 100,000 babies looking at the survivors now instead of those that died. This is the survival curve for the females in 1900. This is the survival curve for females in 1995, and so we asked the question. We know that life expectancy rose by about 30 years during this time period.
Well, if all of the dramatic reductions in death rates observed during the course of the 20th century were to occur again at every age, this is the survival curve that would result. This is the life expectancy that would result, about 89.1. In other words, about a ten year increase in life expectancy at birth, not the 30 years that we observed during the previous century. And if it all happened a third time, the gain would be only 6.1 years, life expectancy would be 95. It's a way to illustrate that the magnitude of the gain and the increase in life expectancy is decelerating as it goes higher. Skip that.
Now I want to present an opposing point of view. Some mathematical demographers have suggested that life expectancy will go to 100 by the year 2060. And this will happen principally by extrapolating past mortality trends into the future. And this is the figure that has been used. And there's very compelling evidence here that there have been some interesting and dramatic increases in life expectancy observed, steady increases since 1840.
This is record life expectancy observed among humans in various sub-groups of the population, increasing steadily from about 45 or so, all the way up to, the record I think now is close to 85 for Japanese females. And what they're suggesting is, is that if we have observed this increase in life expectancy for the past 160 years or so, that there is no reason why it cannot continue into the future. And so this mathematical extrapolation is the basis for the prediction that life expectancy will go to 100 by the year 2060. And if this is all you look at, if you look only at the historical trend in life expectancy, this is a very compelling argument for why you might anticipate a continuation of this historical trend.
Now this is the actual terminology that is used by the authors. They demonstrated that there's been a two and a half year increase in life expectancy per decade, and therefore, it's reasonable to anticipate that this will continue out for the next six decades. However, I used this same extrapolation method to go backward in time just to see what would happen. And if you go backward in time using the same extrapolation method, life expectancy would be zero in the year 1750. So I have argued, as have some others, that it is no more reasonable to make projections of life expectancy going forward in time using an extrapolation method, than it is to go backward in time. We basically need to understand the underlying biology of humans to make sure forecasts, rather than relying entirely on mathematical extrapolations.
So if there's going to be another quantum leap in life expectancy, we're going to have to extend the duration of life, of people like Madame Jeanne Calment and others who have already lived 70, 80, 90 or 100 years or more by 70 or 80 years, in order to achieve a comparable increase in life expectancy like that observed during the 20th Century. And here's where Steve and I are in complete agreement.
The only way this is going to happen, it's not going to happen by altering our lifestyles. It's not going to happen by ingesting anti-oxidants. It's not going to happen by injecting yourself with growth hormone. It's going to have to happen by altering the basic biological rate of aging itself, which is something we cannot currently do, but researchers are trying to do. And there's one last way to illustrate this point.
You know actually, interestingly enough, Steve, this is the first time I've ever seen you show a figure of running times. Here's mine for the world record for the one mile run, indicating that it's declined steadily, very much like life expectancy has increased steadily in the middle of the last century from about five minutes in 1850, to three minutes and 43 seconds today. So if you do a linear extrapolation of this trend, which is the same method that is being used to generate these much higher life expectancies, you would run one minute -- one mile in one minute in the year 2420, and we would do it instantaneously in the year 2580.
Okay. Are there hard physiological barriers to the human life cycle? Now the answer to this question is yes and no, and I'm going to spend much less time on this than I did on the issue of life expectancy, but the yes and no answer, you know, when you say yes and no, some people will only hear the yes, and some people will only hear the no. And I will encourage you to hear both answers, and here's where I get into a bit of evolution biology, which I thought Steve was going to talk more about, but I'll touch upon this issue a bit more.
And it's the issue of why not immortality? Why aren't we immortal? And the answer is, immortality in a way already exists for DNA. And once DNA acquired the property of immortality, its carriers became mortal -- you and I.
Now I use this analogy, and have used it for quite some time as a way to illustrate the fairly complex evolutionary theory of senescence. This is the Indianapolis 500 race car analogy. Now we know, of course, what the duration of this race is. It's 500 miles. If something goes wrong with these automobiles during the race, they bring them in, they fix the parts, they send them back out. What's interesting here is that when the race is over, they turn the engine off and they bring the car back into the shop, which is something we can't do in humans. The engine of life is always operating until the end.
If you were to conduct a hypothetical experiment on Indianapolis 500 race cars, where instead of turning the engine off after 500 miles, you continue to operate them, run them around the track until they all failed, you would actually see a distribution of failure times that is very much like that of living organisms, life humans, and mice, and dogs.
The key thing here is that, number one, you would get to see things go wrong with these automobiles that you would never ordinarily have an opportunity to see, because you're operating them beyond the end of their effective warranty period. And the other key point is, is that the engineers in this case did not build in a program for failure. They're simply operating beyond the time period that they were intended to be used.
Well, the same logic actually applies to sexually reproducing species, but the measure -- the end of the race is not a measure of distance, it is a measure of time. And in this case, reproductive success includes not just the time period when we are producing offspring, but a time period when we can contribute to the reproductive success or fitness of our offspring as grandparents, so there's a grandparenting period as part of the end of the reproductive period. And in effect, what we are doing to ourselves and other sexually reproducing species is the very experiment that I was talking about with automobiles. We are pushing ourselves well beyond the end of our reproductive window, and we are having an opportunity to see things go wrong with our living machines that we never would ordinarily have an opportunity to see. And the further we push the envelope of survival into the post reproductive region of the life span, the more things we will see go wrong, as well.
Now this is the basic linkage that I know Dr. Kass was looking for on this linkage between reproduction and senescence. The basic evolutionary argument suggesting that natural selection is very effective at influencing gene frequencies in the pre-reproductive period, but as soon as we begin reproducing the ability of natural selection to alter gene frequencies declines very rapidly, as soon as we begin reproduction. The force of selection declines very rapidly, until we reach the post reproductive region of the life span, where the ability of selection to influence gene frequencies declines to very low or negligible levels.
Now this actually is a key figure linking reproductive and senescence. And what I will tell you is, and my colleagues and I have done work in this areas, as well, where have looked at the fundamental linkage between the timing of reproduction when puberty begins, the length of the reproductive window, and the duration of life of a species. And a number of researchers have demonstrated that the duration of life of a species is calibrated to the onset and length of the reproductive window. And we've done this in more than a dozen mouse strains from the Argonne National Laboratory database. We found something similar in humans, in dogs, that there is a fundamental linkage between these biological attributes. And if you're going to alter one, you are likely going to alter the other.
I'm going to skip over -- well, actually let me show you. This actually is an important figure. You know, when we talk about death rates, this is what they actually look like. Age on the X axis. This is a semi-log scale, so when there's exponential increases in mortality, you get a straight line here. This is a way to illustrate that there have been -- that the age trajectory of mortality really has not changed very much during the course of the 20th Century, even though we have lived much longer. And this actually is an important point, because many of the researchers from the biological sciences who have suggested and have argued that we have altered aging, I believe, have not actually demonstrated that aging itself has been altered, because we can't measure biological aging itself. It has been used as this, the change in the death rate, both the inflection point, the timing with -- the age at which the death rate increases, and then the slope of this mortality curve, which is taken as a proxy for aging.
The biological process of aging itself cannot currently be measured, so when somebody says that aging has been altered, or aging has been delayed or postponed, I would say that we don't have definitive evidence to support this particular view. This is what is used, so much of what goes wrong with us as we grow older, as many of us have suggested is not our fault. And senescence is, indeed, an accident of surviving beyond the warranty period for living machines, which is the main point we were making earlier.
Now here's the yes and the no. There is no biological limit to life. Evolution could not have given rise to genes designed for the purpose of killing us, but nevertheless, duration of life is fundamentally influenced by biological clocks that regulate growth, development, and reproduction, and senescence is an inadvertent bi-product of these genetic programs according to evolution theory. So on the one hand no, we don't have a program designed to kill us, but on the other hand we do have programs, very tightly controlled genetic programs for growth, development and reproduction, that have as an inadvertent bi-product of their existence, senescence. And so there can be no aging or longevity genes, nor are there hard physiological barriers to extending the life span, but nevertheless, there are constraints on duration of life that are influenced by both biochemical changes that occur, biomechanical changes. This is an article we published last year of human rebuilt to last in biomechanics, then biodemographic constraints. And this is smaller, but I couldn't find a symbol for stochastic events, but it doesn't mean it's any less important, the stochastic or random nature to the aging process itself.
Now I'm going to spend much less time on this issue, Prospects for Significant Breakthroughs in Aging Research. It is important to distinguish between breakthroughs that modify the biological rate of aging, and breakthroughs that may extend the duration of life. And I think some of these animal models are really modifying the duration of life. Much as we do with humans in altering our risk of heart disease and cancer, one could make the argument that we're altering aging. I don't think we are.
Extension of life can occur without modifying the biological rate of aging. If I was asked to come up with any sort of breakthrough that I would anticipate that would modify the biological rate of aging, I would expect it would come from pharmaceutical industry that alters maintenance and repair functions. I put in slightly smaller letters caloric restriction, emetics, and genetic engineering. I'm less convinced that these will actually work on humans, and we can discuss this later.
Breakthroughs that might extend duration of life without modifying the biological rate of aging, there are plenty of these. There's a whole laundry list of ways in which we can intervene that may extend duration of life without necessarily influencing the biological rate of aging. I know you have personal interest in many of these topics.
All right. Demographic implications of life extension, I'll address very quickly. I wasn't sure which one of these -- these are both the same figure, and this one is better, so I'm going to use this one. You know, this is a typical question that students of demography will ask, and so -- and I'm not the first one to answer this. It was actually answered by Anthony Cole, a famous economist in the 1950s at Princeton. And so I did some basic calculations to demonstrate what would happen if we achieved immortality today. And I compared it with growth rates for the population in the middle of the 20th Century. This is an estimate of the birth rate and the death rate in the year 1000, birth rate roughly 70, death rate about 69.5. Remember when there's a growth rate of 1 percent, very much like your money, a growth rate of 1 percent leads to a doubling time at about 69 to 70 years. It's the same thing with humans. With a 1 percent growth rate, the population doubles in about 69 years. If you have the growth rate -- if you double the growth rate, you have the time it takes for the population to double, so it's nothing more than the difference between the birth rate and the death rate to generate the growth rate. And here you can see in 1900, the growth rate was about 2 percent, which meant the doubling time was about five years. During the 1950s at the height of the baby boom, the growth rate was about 3 percent, which means the doubling time was about 26 years. In the year 2000, we have birth rates of about 15 per thousand, deaths of about 10 per thousand, low mortality populations, which means the growth rate is about one half of 1 percent, which means it would take about 140 years for the population to double.
Well, if we achieved immortality today, in other words, if the death rate went down to zero, then the growth rate would be defined by the birth rate. The birth rate would be about 15 per thousand, which means the doubling time would be 53 years, and more realistically, if we achieved immortality, we might anticipate a reduction in the birth rate to roughly ten per thousand, in which case the doubling time would be about 80 years. The bottom line is, is that if we achieved immortality today, the growth rate of the population would be less than what we observed during the post World War II baby boom.
We would eventually run into problems, of course, a century down the road, but just so you know the growth rates would not be nearly what they were in the post World War II era, even with immortality today. However, it would have a rather dramatic effect on age entitlement programs like Social Security and Medicare. When Social Security was created in 1935, they predicted there would be no more than about 20 million beneficiaries. This is what was actually observed, and their recent prediction, if indeed life expectancy were to go much higher, then we would run into very severe problems with the funding of age entitlement programs like Social Security and Medicare.
If there was another quantum leap in life expectancy, I don't really know how these other attributes of human life would change, marriage, retirement, work, education. I don't know any better than anyone else. What I would anticipate is, is that there would be fundamental differences in all of these attributes of society if there was another quantum leap in life expectancy.
I'm going to end with the position statement on human aging, and I'm only going to touch upon this briefly. This is something we published last year. You might imagine getting 51 scientists to agree on anything is extraordinarily difficult. In fact, as hard as I tried to get Steve to sign onto our position statement, we weren't able to get him to sign on. I will tell you that we didn't all agree on everything that we put into this position statement, but we decided to compromise on some of the language for the purpose of getting across a very important message to the public. And that is, number one, there are no anti-aging medicines in existence today. And number two, there is a great deal of very interesting, fascinating good research ongoing in the field of gerontology designed to understand and modify the biological rate of aging. And we not only support this research, but believe that it's absolutely critical to helping us deal with a much more rapidly aging population. And there was a question associated with this, do the conclusions that we came to in our position statement apply to biomedical interventions? Let me skip -- well, actually, I don't want to skip by this.
A number of other publications came out associated with this question of anti-aging medicine. This one came out from International Longevity Center in New York. This is a report that was published by the GAO the week of 9/11, which is why no one saw it, suggesting that anti-aging products pose a potential for physical and economic harm. And I will end by showing you the various issues that we raised during this discussion. I mean, really this position statement grew out of a AAAS meeting that many of us attended about a year and a half ago, where we were discussing for prospects for increasing human life expectancy, and we were lamenting about the problems associated with those selling anti-aging products, so we decided to provide as definitive a statement as we could about each of these issues, about what we know and what we don't know, and whether or not we think that these influence aging itself or duration of life. And so we basically have -- we tried to create as short a paragraph as we could on each one of these issues to tell you what we knew, and what we didn't know. And I'm going to end with, I know you don't want to read any of this stuff. And we have published this, but I will point out the very last sentence of our conclusion, which is, for those of you who cannot read this, "Successful efforts to slow the rate of aging would certainly have dramatic health benefits to the population by far exceeding the anticipated changes in health and length of life that would result from the complete elimination of heart disease, cancer stroke, and other age-associated diseases and disorders." So I completely agree with Steve on this point, which is one of the first points that he was making; and that is, is that research on aging is fundamental. If we can succeed in postponing many of the diseases and disorders associated with aging, the benefits would be far exceed those that would accrue, we believe, with the elimination of major fatal diseases in the population. And I think I'll end there.
CHAIRMAN KASS: Thank you very much. We'll get the lights on shortly, but if people are willing to start in semi-darkness, I think I see Robby George's hand. Please.
PROF. GEORGE: Thank you, Dr. Olshansky, for that presentation. One of the figures you had on death rates included, if I saw it correctly, a rather startling statistic on an increase in death rates for persons between the ages of 20 and 35 from the period of 1985 forward. Do you remember that figure? For the U.S., yeah.
CHAIRMAN KASS: You commented on it.
DR. OLSHANSKY: Was it this one?
PROF. GEORGE: No, I don't believe so.
CHAIRMAN KASS: It was the table that you showed with Japan and France.
DR. OLSHANSKY: Oh, this one.
PROF. GEORGE: Yeah. What's the plus one point? I mean, that jumps off the page.
DR. OLSHANSKY: Yes, it should jump off the page.
PROF. GEORGE: 20 to 39, I thought it was 35. It's 39.
DR. OLSHANSKY: 20 to 39, and it's actually -- I didn't show this, but there are increases in death rates among some older age groups, individuals at older ages, as well. That's a plus 0.7 percent increase, seven-tenths of 1 percent, a 1.1 percent increase for males probably associated with HIV. But yes, it should jump off the page.
It's a way to illustrate, I think, that -- you know, there's a tendency when making forecasts of life expectancy to assume that they're always going to rise. And this issue came up when I was here in Washington a couple of months ago talking to the Trustees of the Social Security Administration about this very issue, about projections of life expectancy. And it is always assumed that they are going to go up, and I have suggested, as have others, that maybe we need to be cautious about this long term projection assuming it's always going to rise. There are sub-groups of the human population that have experienced an actual decline in life expectancy in various parts of the world. Also, at some age groups that there have been increases in mortality, not always decreases in mortality, so you're correct.
DR. ROWLEY: And what's the role of guns and killing with guns in that particular age group?
DR. OLSHANSKY: Well, that's the age window where you would see the effect of extrinsic causes, like homicide and accidents. And clearly, if those are on the rise, you're going to see an increase in mortality in those age groups.
PROF. GEORGE: Do you know if they are on the rise for that period?
DR. OLSHANSKY: Well, for that period they were on the rise, absolutely. This is 1985 to 1995.
PROF. GEORGE: I'm asking whether you know if the murder rate is on the rise.
DR. OLSHANSKY: Oh, I don't know. I didn't break down the seven-tenths of a year, or the 7 percent and the 1.1 percent into the underlying causes. I'm guessing that HIV has contributed significantly to this in the United States, but I don't know about the change in homicide rate.
PROF. GEORGE: But if it were HIV, should the statistics be that out of whack with what's going on in France? I just don't know. We're talking about a rate here.
DR. OLSHANSKY: This is a percentage change in the conditional probability of death, and I just don't -- I didn't break it down, so I don't know with certainty what led to the increase in mortality in that age range. But those are real numbers for the United States.
DR. ROWLEY: But isn't homicide the most common cause of death in males, particularly black males, age something like 18 to 30?
DR. OLSHANSKY: I think so, yes. Yes.
CHAIRMAN KASS: Dan Foster.
DR. FOSTER: There are certain precincts in Washington where the life expectancy of young males in certain racial groups is, y
