The Italian study argues after a certain threshold age, lets say 105, Italians hit a constant, "mortality plateau, which was 50:50 odds of living to your next birthday, and thus since mortality is not increasing year on year, theoretically there is no upper bound on the age of a person. This may be true theoretically, but mathematical reasoning clarifies this somewhat. Even if you have 1 million humans aged 105 right now, which is probably a vast overstatement, they would diminish to 100,000 humans in less than 4 years, then diminishing to 1,000 humans aged 115 in 10 years' time. After this point, it's basically a question of statistics, as we keep halving the population yearly and get into statistically unsubstantial cohorts of humans. So, while it may be that some humans could live to 130-140 naturally, without fountain-of-youth innovations, the VAST majority of us will die much, much earlier.
The low-hanging fruits of obesity, heart disease, and diabetes are calling to public health far more urgently than this pie in the sky limitless life.
> The low-hanging fruits of obesity, heart disease, and diabetes are calling to public health far more urgently than this pie in the sky limitless life.
Still, it's definitely worth it to continue research along these lines and plenty of other ridiculous lines. Even if they don't come to anything. We have such a tremendous abundance of intellectual capacity as a species. If we dedicate all our intellectual capacity to solving the low-hanging fruits in healthcare (or to increasing the click-through rates of ads, among other things) it would require us to allocate many people to no task at all, just because there was already an excessive abundance of resources already dedicated to the tasks.
It's a little fun and encouraging to think about, but the more people there are alive, the more important unimportant things become, because with more people there are more problem solvers.
I really appreciate this perspective. More people is fundamentally a good thing. Some would vehemently argue otherwise, but I believe the problems they bring up come down to a calculus that depends first on other things, like profit vs. environmental stewardship, etc., not primarily the birth and death rates.
> mortality plateau, which was 50:50 odds of living to your next birthday
Isn't the question whether or not we can affect the odds here? If we can increase it to 75:25 at this plateau, then we also are also increasing the age longest lived of these people are predicted to live for immensely, and the last person of this example group of 1 million starting at age 105 you give would expect that last person to die not 10 years later (at 115) on average as with a 50:50 mortality, but around 48 years later, at 153. Every slight change in that mortality rate affects that long tail quite a bit, relatively speaking.
I for one believe that might have an immense effect on how people view their health and longevity. If people believe that taking care of themselves might give them many decades of additional life instead of a few years, that could lead to a wider shift.
> The low-hanging fruits of obesity, heart disease, and diabetes are calling to public health far more urgently than this pie in the sky limitless life.
Vastly more people die each year of age-related causes than obesity, heart disease, and diabetes combined. It's an unrecognized holocaust inevitably terminating the potential of all humans on this planet, and it is going entirely unrecognized because if you work on solving aging you are a crackpot, whereas "obviously" obesity, heart disease, and diabetes are more deserving of our time and grant money.
What is your view of the lives of elderly people? I would trade health, mobility and my intellectual capabilities intact to 90 for 20 years of problems (70->90) followed by another 20 years of geriatric life (90->110).
I would agree with you if the common experience of old age was positive and healthy; but it seems to me that the priority for research should be to alleviate the problems and suffering we will all face should we be lucky enough to get there, especially now that the majority of people are lucky.
Perhaps healthier older people will live a lot longer as well?
Humans have a 1/3 lifetime cancer risk, and I suspect it will go up if we extend lifetimes. It's a wall to longevity. If we get past that wall, there's still proteopathys. There's no real approach to curing those yet (at least not that I'm aware of.)
>Humans have a 1/3 lifetime cancer risk, and I suspect it will go up if we extend lifetimes. It's a wall to longevity.
No, it's not. Cancer is just a form of malfunction of cellular machinery, and as such, can certainly be mitigated or cured at some point. There's already lots of promising research in eliminating it, such as by re-purposing viruses to program the body's immune system to recognize and eliminate the cancerous cells.
It might require artificial intervention, but I believe cancer will be mostly eliminated before too long. And artificial intervention shouldn't be a problem: we have to do this on all our other machines too, we just call it "repair" or "maintenance". There's no such thing as a machine which never needs maintenance from an external source.
Why do you think cancer is undefeatable? The risk of cancer rises substantially with age, and much of the increase in risk is probably due to the immune system malfunctioning.
They also agree with that, but they use a model that can never reach 0. So starting with nearly 4000 105yo individuals, only one remains at 115 years and never dies.
So biologists say there is a limit. But if you only look at the statistics from extreme examples of people over 105, mortality looks flat for a few years, so statisticians say we can live forever.
It's not bad logic, but it is a non-obvious conclusion. Which is why it gets misreported.
One of two things may be true:
1) As we age, mortality rate increases until certain
2) As we age, mortality rate increases, and then levels off at some point
The latter doesn't mean you'll live forever: you're still already rolling the dice at the highest handicap every year. But it does mean there's no fundamental age everyone MUST die by (aka "live forever", if you're one in a billion).
The more interesting differentiation between the two, and the reason biology is interested in the question, is as an answer to "If we want to maximally prolong life, what do we do?"
If the flattening risk (option 2) model is correct, then we could extend life very far indeed by simply preventing people from dying of things that normally kill them by age 85 (e.g. disease, cancer) and then provide them with high quality medical support (to best fight the risk of their expiring from "normal" causes, e.g. accidents).
On the other hand, if it's ever-increasing risk (option 1), we should instead spend our effort at untangling the microbiology mysteries at the root of aging, and see if we can instead re-engineer systems to not work the way they currently do.
Or as my father (PhD in pathology) quips on the whole matter, "At some point, old folks are past the point that all of things that normally kill people would have killed them. And by virtue of still being alive, it's probable they're more resistant. We say they've reached escape velocity."
Imagine you keep flipping 10 coins, stopping each once it has come up heads. Eventually all will have come up heads. But if you do this a bunch of times and keep track of how many times you flipped the last coin, there will be a certain number of flips at which the end is most commonly reached.
The "mortality rate" increasing then flattening out is just this curve reaching its peak then declining. The math of this is easy enough to figure out so not sure why there is a mystery amongst biologists about it.
People have a 50% chance of dying each year over 105 translates to there is no limit to longevity?
Also, the critic in me makes me want to see the actual distribution because I would think there is a tail effect going on and a small group of people are just genetically living much longer.
Death rates increase every year due to naturally occurring mechanisms in the body. Consider those mechanisms as a function with the body as a parameter and the output as a more aged body (or rather a body with higher probability of failure).
If the body reaches a point where the death rate ceases to increase year per year, it would suggest that those same naturally occuring mechanisms, thought of as a function, approaches some asymptote.
Note this doesn't necessarily mean you stop aging (in appearance), just that your body's rate of failure no longer increases. This suggests that if the reasons for failure can be treated properly, perhaps the body can remain in that equilibrium state indefinitely.
This study provides some preliminary evidence that such an asymptote may exist in the body's natural mechanisms. Both controversial and exciting. Worth exploring further.
More "journalism" with no understanding of numbers? That can't be. (Sarcasm). Just the same, this is also the same profession that still doesn't know the difference between cause and correlation. How can you be allowed to graduate with a degree in journalism and not understand cause v correlation?
That being said, it seems that reading just about an scientific study these days makes you wonder if those doing the publishing understand data analysis, statistics, etc. It seems that too often they find some date, retro-fit a conclusion, ignore the flaws and holes, but still publish because they know the "journalists" won't figure it out either.
And then science complains about the public not buying into ever word science mutters? I don't get it.
It means that they found data not agreeing with the model of limitless increase of mortality with age. This is just a statistical observation, not physics yet, not by a long shot.
If the statistical claim is true, that morality rates don't increase after 105, then it is "suggestive of the idea" that there isn't a structural limit to age, even if this doesn't in any way yield some immediate immortality (notice how weakly I phrase that). If the chance of dying is the same at 105 and 115, then maybe the things killing one at 105 and at 115 have the same root cause and heading them off could significantly increase lifespans. The alternative scenario is new things appear that would kill one at 115 but not at 105 but one would naively assume that the mortality increases (it may be old things retreat and new things appear however).
Obviously, there are a lot of "ifs" here but the result seems interesting and suggestive.
A 50% mortality rate per year translates into there being a <0.1% chance of living for 10 years. How does that translate into there being "no limit to longevity?" I seriously don't get the claim.
It's not a valid claim IMHO.
If you look at the sample size that they're basing it on, you can see that it's shrinking at the attrition rate for what they are measuring.
Once you get to 117 years old, there are only a few people worldwide that can be used to determine that survival rate, and those people are already subject to a selection bias of people that are exceedingly long-lived.
As we live, we forget stuff, we change, the situation around us changes. So it's as if we gradually die every day, to be replaced with a new self. The day we physically die is just the last time we die. Not to mention that we experience nothingness and coming back every night.
By analogy, is a continually trained neural net the same as a previous version of itself? Is it a different one, especially after a long time? Did the old version end/die by continuous accumulation of changes?
Contrast your idea to the ideas used in the search for life on Mars. The Martian atmosphere is very near its chemical equilibrium, the Earths atmosphere is dynamic; Mars is (more or less) dead, Earth is (not as) alive (as it should be if we weren't messing it up). My point is that change is life; stasis is death. On the day we die our body begins a journey to chemical equilibrium, our brains reach electrical equilibrium rather quickly. Then they stop there.
This was the insight of the pre-Socratic philosopher Heraclitus: “No man ever steps in the same river twice, for it's not the same river and he's not the same man.”
I would have thought it's intuitively obvious that there's "no fixed limit on human longevity", and it's therefore obvious that mortality "flattens out": how else could it avoid going over 100%? It presumably is possible to say something interesting and non-trivial about the shape of the curve but it seems to be beyond the mathematical competence of the people writing these journalistic reports to do so.
Looking at the Wikipedia list of longest-lived people (https://en.wikipedia.org/wiki/List_of_the_verified_oldest_pe...) I get the impression that the top two (Jeanne Calment and Sarah Knauss) don't fit, though I know that the case of Jeanne Calment is very well documented.
It's not intuitively obvious at all, since each replication of our cells leads to our chromosomes getting ever smaller until presumably it starts eating into genes that do useful things. Unless it's intuitively obvious to you that you could live without genetic material at all, it seems quite obvious that -- given current biological processes -- there is a limit, even if that limit is longer than where we've gone before.
Even if it were the case that each replication made the chromosomes smaller (personally I would guess that there is a non-zero probability of them not getting smaller and even a non-zero probability of them getting larger thanks to a mutation) that would still only imply a hard limit on the number of replications. You would also need to find a hard limit for how long a cell can survive without replicating. Good luck with that.
I content that there are no limits: however old you are, you have a non-zero chance of surviving another picosecond. The product of a finite number of non-zero real numbers is non-zero, therefore...
Some people may wish to retort that at some point these numbers become so small that they are no different from zero in practice. Those people are probably not mathematicians. They're not wrong, either, of course, but they're not making a valid argument for there being a limit.
There's no limit for the number of successive heads you can get when tossing a coin, either, though you won't in practice see more than a hundred if it's a reasonably fair coin. If there's no limit in that simplified, mathematical situation, does it make sense to hypothesise that "there is a limit" in the case of highly complex biochemical processes?
This is all very philosophical. It doesn't help us to estimate the chances of a given person who is alive and 115 years old today (there seem to be about five of them) one day breaking Jeanne Calment's record of 122. Are any bookmakers taking bets?
There's the natural limit and the theoretical limit. In theory, some future gene therapy may swap out your chromosomes for ones that don't decay (while somehow not causing negative side effects) while you're still an embryo or gamete, and do all sorts of other things to make you nearly biologically immortal. And in some even more distant future, maybe consciousness upload (with the "copy conundrum" somehow resolved) will be a real thing, in which case living for thousands or millions of years seems plausible.
Even if there were a hard-coded expiration mechanism, that mechanism would presumably have limited precision and a non-zero failure rate (assuming it's implemented with biochemistry rather than magic). Looking at the list I linked to, it does rather look as if there were a cunning device to kill people towards the end of their 118th year, and that device failed in two cases... but I can't imagine why there should be such a device or how it might be implemented, so I don't really believe that.
Even if the chances of surviving another year did stabilise around 50% after the age of 100, say, you still wouldn't have enough people living beyond 130 for you to be able to measure with reasonable accuracy the mortality rate of people aged over 130. So the hypothesis that a person aged 140 would still have a 50% chance of surviving another year would be totally untestable.
If you think about it, there's also no hard limit for how tall or how short an adult human could be. Except the limit of zero for the height. Again, beyond a certain point there will be, in practice, no data available for testing any hypothesis about the distribution.
In the terms of probability aging distribution becomes memoryless for very old people. This would mean that the remaining lifespan for really old people would follow exponential distribution. Just like with radioactive decay, you could determine their half life. The half life could be as short as one year.
>At that point, the researchers say, the odds of someone dying from one birthday to the next are roughly 50:50
This really has nothing to do with limits to longevity in any practical sense. That's just the publicity machine at work, the first line in mangling and misrepresenting the relevance of research. This paper is to do with a long-running debate over whether or not mortality risk reaches a high, flat plateau in extreme old age. The difference between limit and high flat plateau is academic for those involved. The outcome is much the same.
It does in flies, and there is very clear evidence for that. The evidence in humans to date is very sparse, and where people have done good work on the numbers, they have not seen signs of a plateau. So in that sense, that people have found a larger data set that shows it, that is fuel for the fire.
No-one has the slightest idea as to how you can balance the fundamentals of reliability theory with a late life mortality plateau. If damage is accumulating, and no-one really believes it stops accumulating, then your risk of mortality should keep on going up. What sort of mechanisms in a highly damaged biology could buffer that underlying reality? And why in very late life versus other times?
None of this, of course, has anything to do with efforts to treat aging or extend human life. It is an academic debate that will likely fade out before answers are found, made irrelevant by the advent of ways to treat aging and its causes. Methods of extending life will work by putting off or repairing the very damaged state. There will likely never be a large multiple of the present count of exceptionally physiologically aged individuals. Rejuvenation therapies are just around the corner.
I wonder if there is something like the study I heard of once that said, if an electronic or mechanical device lasts at least X years, it will almost always last X + Y years where Y is a much longer period of time. I heard that related to design, manufacturing process and material selection.
If it's true for things, it makes sense it might be true for humans.
Readit News
The low-hanging fruits of obesity, heart disease, and diabetes are calling to public health far more urgently than this pie in the sky limitless life.
Still, it's definitely worth it to continue research along these lines and plenty of other ridiculous lines. Even if they don't come to anything. We have such a tremendous abundance of intellectual capacity as a species. If we dedicate all our intellectual capacity to solving the low-hanging fruits in healthcare (or to increasing the click-through rates of ads, among other things) it would require us to allocate many people to no task at all, just because there was already an excessive abundance of resources already dedicated to the tasks.
It's a little fun and encouraging to think about, but the more people there are alive, the more important unimportant things become, because with more people there are more problem solvers.
I really appreciate this perspective. More people is fundamentally a good thing. Some would vehemently argue otherwise, but I believe the problems they bring up come down to a calculus that depends first on other things, like profit vs. environmental stewardship, etc., not primarily the birth and death rates.
Isn't the question whether or not we can affect the odds here? If we can increase it to 75:25 at this plateau, then we also are also increasing the age longest lived of these people are predicted to live for immensely, and the last person of this example group of 1 million starting at age 105 you give would expect that last person to die not 10 years later (at 115) on average as with a 50:50 mortality, but around 48 years later, at 153. Every slight change in that mortality rate affects that long tail quite a bit, relatively speaking.
I for one believe that might have an immense effect on how people view their health and longevity. If people believe that taking care of themselves might give them many decades of additional life instead of a few years, that could lead to a wider shift.
Please correct me if my math is wrong or if I'm miss interpreting how to calculate this.
Vastly more people die each year of age-related causes than obesity, heart disease, and diabetes combined. It's an unrecognized holocaust inevitably terminating the potential of all humans on this planet, and it is going entirely unrecognized because if you work on solving aging you are a crackpot, whereas "obviously" obesity, heart disease, and diabetes are more deserving of our time and grant money.
I would agree with you if the common experience of old age was positive and healthy; but it seems to me that the priority for research should be to alleviate the problems and suffering we will all face should we be lucky enough to get there, especially now that the majority of people are lucky.
Perhaps healthier older people will live a lot longer as well?
No, it's not. Cancer is just a form of malfunction of cellular machinery, and as such, can certainly be mitigated or cured at some point. There's already lots of promising research in eliminating it, such as by re-purposing viruses to program the body's immune system to recognize and eliminate the cancerous cells.
It might require artificial intervention, but I believe cancer will be mostly eliminated before too long. And artificial intervention shouldn't be a problem: we have to do this on all our other machines too, we just call it "repair" or "maintenance". There's no such thing as a machine which never needs maintenance from an external source.
Is it just me, or is this just bad logic?
One of two things may be true:
1) As we age, mortality rate increases until certain
2) As we age, mortality rate increases, and then levels off at some point
The latter doesn't mean you'll live forever: you're still already rolling the dice at the highest handicap every year. But it does mean there's no fundamental age everyone MUST die by (aka "live forever", if you're one in a billion).
The more interesting differentiation between the two, and the reason biology is interested in the question, is as an answer to "If we want to maximally prolong life, what do we do?"
If the flattening risk (option 2) model is correct, then we could extend life very far indeed by simply preventing people from dying of things that normally kill them by age 85 (e.g. disease, cancer) and then provide them with high quality medical support (to best fight the risk of their expiring from "normal" causes, e.g. accidents).
On the other hand, if it's ever-increasing risk (option 1), we should instead spend our effort at untangling the microbiology mysteries at the root of aging, and see if we can instead re-engineer systems to not work the way they currently do.
Or as my father (PhD in pathology) quips on the whole matter, "At some point, old folks are past the point that all of things that normally kill people would have killed them. And by virtue of still being alive, it's probable they're more resistant. We say they've reached escape velocity."
The "mortality rate" increasing then flattening out is just this curve reaching its peak then declining. The math of this is easy enough to figure out so not sure why there is a mystery amongst biologists about it.
People have a 50% chance of dying each year over 105 translates to there is no limit to longevity?
Also, the critic in me makes me want to see the actual distribution because I would think there is a tail effect going on and a small group of people are just genetically living much longer.
Death rates increase every year due to naturally occurring mechanisms in the body. Consider those mechanisms as a function with the body as a parameter and the output as a more aged body (or rather a body with higher probability of failure).
If the body reaches a point where the death rate ceases to increase year per year, it would suggest that those same naturally occuring mechanisms, thought of as a function, approaches some asymptote.
Note this doesn't necessarily mean you stop aging (in appearance), just that your body's rate of failure no longer increases. This suggests that if the reasons for failure can be treated properly, perhaps the body can remain in that equilibrium state indefinitely.
This study provides some preliminary evidence that such an asymptote may exist in the body's natural mechanisms. Both controversial and exciting. Worth exploring further.
That being said, it seems that reading just about an scientific study these days makes you wonder if those doing the publishing understand data analysis, statistics, etc. It seems that too often they find some date, retro-fit a conclusion, ignore the flaws and holes, but still publish because they know the "journalists" won't figure it out either.
And then science complains about the public not buying into ever word science mutters? I don't get it.
It just sounds dumb because it doesn't take into account any physical limitations that would make the curve reach zero.
Obviously, there are a lot of "ifs" here but the result seems interesting and suggestive.
Once you get to 117 years old, there are only a few people worldwide that can be used to determine that survival rate, and those people are already subject to a selection bias of people that are exceedingly long-lived.
By analogy, is a continually trained neural net the same as a previous version of itself? Is it a different one, especially after a long time? Did the old version end/die by continuous accumulation of changes?
Looking at the Wikipedia list of longest-lived people (https://en.wikipedia.org/wiki/List_of_the_verified_oldest_pe...) I get the impression that the top two (Jeanne Calment and Sarah Knauss) don't fit, though I know that the case of Jeanne Calment is very well documented.
I content that there are no limits: however old you are, you have a non-zero chance of surviving another picosecond. The product of a finite number of non-zero real numbers is non-zero, therefore...
Some people may wish to retort that at some point these numbers become so small that they are no different from zero in practice. Those people are probably not mathematicians. They're not wrong, either, of course, but they're not making a valid argument for there being a limit.
There's no limit for the number of successive heads you can get when tossing a coin, either, though you won't in practice see more than a hundred if it's a reasonably fair coin. If there's no limit in that simplified, mathematical situation, does it make sense to hypothesise that "there is a limit" in the case of highly complex biochemical processes?
This is all very philosophical. It doesn't help us to estimate the chances of a given person who is alive and 115 years old today (there seem to be about five of them) one day breaking Jeanne Calment's record of 122. Are any bookmakers taking bets?
Even if the chances of surviving another year did stabilise around 50% after the age of 100, say, you still wouldn't have enough people living beyond 130 for you to be able to measure with reasonable accuracy the mortality rate of people aged over 130. So the hypothesis that a person aged 140 would still have a 50% chance of surviving another year would be totally untestable.
If you think about it, there's also no hard limit for how tall or how short an adult human could be. Except the limit of zero for the height. Again, beyond a certain point there will be, in practice, no data available for testing any hypothesis about the distribution.
In the terms of probability aging distribution becomes memoryless for very old people. This would mean that the remaining lifespan for really old people would follow exponential distribution. Just like with radioactive decay, you could determine their half life. The half life could be as short as one year.
>At that point, the researchers say, the odds of someone dying from one birthday to the next are roughly 50:50
It does in flies, and there is very clear evidence for that. The evidence in humans to date is very sparse, and where people have done good work on the numbers, they have not seen signs of a plateau. So in that sense, that people have found a larger data set that shows it, that is fuel for the fire.
No-one has the slightest idea as to how you can balance the fundamentals of reliability theory with a late life mortality plateau. If damage is accumulating, and no-one really believes it stops accumulating, then your risk of mortality should keep on going up. What sort of mechanisms in a highly damaged biology could buffer that underlying reality? And why in very late life versus other times?
None of this, of course, has anything to do with efforts to treat aging or extend human life. It is an academic debate that will likely fade out before answers are found, made irrelevant by the advent of ways to treat aging and its causes. Methods of extending life will work by putting off or repairing the very damaged state. There will likely never be a large multiple of the present count of exceptionally physiologically aged individuals. Rejuvenation therapies are just around the corner.
If it's true for things, it makes sense it might be true for humans.
https://en.wikipedia.org/wiki/Bathtub_curve