Will Our Average Life Expectancy Ever Reach 100?

It’s a inevitable fact of life: Everyone dies. But thanks largely to major medical advances, improved sanitation, and better environmental protections, our collective life expectancy has steadily risen year after year starting in the 19th century (outside of a few pandemics). More recent data, however, has suggested that this gravy train is grinding to a crawl.

A study published this past October in Nature Aging, for instance, found that, while life expectancy in countries like the U.S. and other high-income nations has continued to grow, the rate of increase has substantially slowed down over the past thirty years. Moreover, the odds of someone reaching the age of 100 these days remain very low; in the U.S., only 3.1% of women and 1.3% of men born in 2019 are expected to become a centenarian. In light of the findings, the researchers behind the study argue that humanity has started to brush up against the barriers of our natural mortality, and that our current approach to treating age-related diseases like cancer will likely only provide diminishing returns and incremental boosts in longevity moving forward.

Not everyone is quite so pessimistic about the future of aging, though. For this Giz Asks, we asked several experts whether the average person might one day reach 100—and, more generally, if there’s a hard limit to human longevity and how we might break through it.

S. Jay Olshansky

Professor of epidemiology and biostatistics at the University of Illinois at Chicago’s School of Public Health; longevity researcher; and lead author of the Nature Aging paper.

Well, forever is a very long time, so I can’t answer this using an indefinite time frame. What I can say with confidence is that a life expectancy at birth of 100 cannot happen in this century for any national population. In fact, it’s unlikely for most people alive today. Why am I so confident in this answer?

The demographic metric of life expectancy is based on death rates at all ages. The first revolution in longevity occurred because public health saved the young from dying from communicable diseases. Life expectancy experienced a meteoric rise in the 20th century because decades of life were added to the lives of children, but this source of improvement can only happen once for a population—and it’s already been accomplished. Those saved from dying early now have the privilege of living a long life, and humanity should be grateful for the gift of long life afforded to us. But now the currently immutable force of biological aging gets in the way. As long as the aging process is immutable, and larger segments of each birth cohort are exposed to this immutable force of mortality, gains in life expectancy must decelerate. Our latest paper documents the fact that this phenomenon has been happening for the last three decades—so the evidence is not speculative, it’s definitive.

The real question here is what might the future bring? Let’s get one issue addressed immediately. Is it possible that treating diseases as we do now—one at a time as if independent of each other—will allow for a life expectancy at birth of 100? The answer is a definitive no! We demonstrated in 1990 that the virtual elimination of all major fatal diseases will not make humanity immortal—it won’t even lead to a life expectancy that reaches 100. Human longevity is driven by competing risks at advanced ages, so like a game of whack-a-mole, reduce one disease, and two more pop up shortly thereafter.

What about the development of gerotherapeutics [the field focused on developing treatments to slow aging and extend healthy lifespan] that may slow aging? I’m personally excited at the prospect, but the problem here is that a life expectancy at birth of 100 requires that death rates from all causes, at all ages, decline by over 80% from levels present today. Reductions in death rates of this magnitude won’t happen instantly; it takes time to develop and test gerotherapeutics for safety and efficacy; there are serious concerns about inequity in the distribution of such interventions; and it is virtually impossible to empirically demonstrate in a short time window that any gerotherapeutic will yield radical changes in life expectancy. What this means, basically, is that even if science developed an intervention that could make everyone live to 120 or higher, this gain in life expectancy at the population level cannot be proven using the tools of science—even if this hypothetical therapy came online today.

So, is it theoretically possible for a national population to achieve a life expectancy at birth of 100 years? The answer is yes—evolution did not give rise to longevity time bombs that go off at specified ages beyond which humans live today. However, there are also no evolutionary constraints on humans running a mile in one minute, but that is unlikely to happen in these bodies given our current design feature. The same holds true for life expectancy. The only way a life expectancy at birth of 100 is possible is if some new dramatic medical/scientific advance comes online that simultaneously influences every aspect of human aging (both body and mind)—but even if this was happening right before our eyes today, there would be no way to empirically justify that a life expectancy of 100 would be the result.

Nir Barzilai

Geneticist; director of the Institute for Aging Research at Albert Einstein College of Medicine; and lead researcher of the Longevity Genes Project, which has been examining the genetics of over 500 healthy people between the ages of 95 to 112.

One of the questions we wanted to study with centenarians was: Do they get sick when everyone gets sick, starting around the age of 60? Would they just be sick for another 40 years, for instance—which wouldn’t sound so good. But we found the opposite; we found that their longevity and their health went together. They would get diseases between 30 and 50 years after other people got diseases. And it’s not only that they’re living healthy longer, there’s a contraction of morbidity. That means that if they had a disease, it would be for a very short time at the end of their lives. About 30% of them die without taking any drugs and without having any disease. So the idea that humans have the capacity to be healthy for 100 years—are there people like that? Yes. And it’s not a majority of people, but there are hundreds in my study, and probably hundreds of thousands of them around the world.

Now this recent paper seems to be saying, ‘No, you really don’t get over the age of 88.” And look, this paper is written by demographers, and demographers are looking at the past to try to predict the future. These authors are arguing against other demographers who have said: “Hey, look at the last 150 years. Life expectancy has increased linearly, without stop, in a straight line. Every 10 years, we just live 10 years longer, right? And Jay Olshansky has said, ‘Well, is there a roof? Because if there’s a roof, then we’re not going to increase linearly forever.’ And look, I think there’s a roof. There was this Nature paper [published in 2016] that took data and found that the maximal lifespan of humans as a species is around 115 years, That’s not the top for everyone, but that’s the statistical top. If the top is 115 years and now half of us continue to live over the age of 80, there is that roof and it’ll start to bend, because we’re not going up in a straight line and it’s harder to reach that point. But the people in my study seem to have longevity genes that allow them to do that. So we need to harness these genes and have medications for others in order to at least go to that stage.

Would the next stage then be to increase that roof? Is it possible in the future? The answer is probably, I think so. Is it going to be in my lifetime? I don’t think so. Even in animals, we don’t have the proof of concept that you can increase lifespan by so many times. So I’m not saying that this limit wouldn’t be broken, I’m saying something totally else. I’m saying that if we die before the age of 80, but we have 115 years of potential life, we should work on how we can actually get those 35 additional years. How can we live into our 90s or 100s without being sick? That, I think, is more than possible.

The title of the [Nature Aging] paper is about radical life extension. For me, radical life extension is over the age of 150. Is that possible? No, because I think there’s a roof. But can we live longer and better? Certainly.

Preston Estep

Biologist; former chief scientific officer and co-founder of Veritas Genetics; chief scientist and founder of the nonprofit Rapid Deployment Vaccine Collaborative (RaDVac); and entrepreneur in biotech and artificial intelligence research.

The future of human longevity depends entirely on the future of AI. Let’s back up a moment and unpack the reasons why, because some of the more important dependencies on AI might seem remote, but I argue they are far more likely than the commonly imagined scenario of human scientists solving biological aging.

There are two main contributors to advances in longevity: environmental and biomedical. Major advancements in public health and standard of living have driven past gains in average adult life expectancy. Biomedical advancements are now driving impressive gains and promise to increase the average; but whether or not the average will reach 100 years is highly uncertain due to worrisome signs of regression of previous accomplishments. For example, rising anti-vaccine sentiment combined with climate change and other pandemic drivers threaten to further undermine longevity trends that were reversed by the SARS-CoV-2 pandemic. It is more likely that we’ll continue to see a fragmentation of the longevity distribution: the longevity of the wealthiest and most educated will continue to increase, while that of the least educated and anti-science segments of society will lag substantially, and possibly decrease.

As for increasing the longstanding upper limit of human longevity, there is no insurmountable physical barrier; but are radical life extension breakthroughs in biomedical science foreseeable? An instructive example comes from physics. The frontier of physics has been largely stalled for decades, suggesting that even the best human minds have reached an upper cognitive limit. Plus, the number of human-made problems that have the potential to reduce human longevity is increasing. To solve the many extremely complex problems required for radical life extension—and to grow and preserve the infrastructure of civilization necessary to support extreme longevity—human-level intelligence clearly is not up to the task.

Therefore, there is only one possible route to breaking through the upper bound of human longevity: superhuman artificial intelligence. As AI becomes substantially more intelligent than humans, many breakthroughs become possible that are currently impossible, including a substantial increase in the longevity of biological humans. But then some key questions arise. Will biology remain the substrate of choice, or will the fastest and most efficient path to extreme longevity (and superhuman intelligence and power) be to merge with AI?

I argue that merger is the most promising path, in part, because humanity is already merging with AI. We’re not merging as individuals, or even in human form. We’re merging in the form of digital computers and we’re uploading the collective human essence, which has been mined from internet communications and media, and distilled into training data for frontier large language models. The main reason that ChatGPT succeeded wildly after decades of previous failures of generalized AI is that it is the first truly humanized AI. But ChatGPT is just the beginning of a true AI revolution that is attracting massive investments to build bigger and better. Subsequent generations of AI will be more powerful at reasoning, scientific discovery, and the kinds of engineering that will be required for the next steps of merger between humans and AI. And recursively self-improving AI will be able to develop even better technologies for transferring the essence of humans—both collectively and individually—to non-biological computing substrates. That is the most direct path to extreme longevity—maybe even immortality; but it will also enable the production of technological advances capable of producing extreme longevity of biological humans, opening an unprecedented range of choices, trade offs, and dilemmas.