How Long Can You Live?

The oldest people are mostly a paperwork error

This is the part that genuinely changed how I read every longevity headline. The reason we cannot be sure about the wall is that the data at the wall is, in one researcher's words, rotten from the inside out.

The researcher is Saul Justin Newman, and in 2024 he won an Ig Nobel Prize for a deceptively simple discovery: the places and records that produce the world's oldest people line up almost perfectly with the places and records you would expect if the ages were wrong. Newman 2024 Ig Nobel good Not longevity. Poverty, missing documents, and pension fraud.

The cleanest single piece of evidence is what happens when real birth certificates arrive. In the United States, when a state started issuing them, the number of people claiming to be 110 or older fell by 69 to 82 percent. The "supercentenarians" were mostly an artifact of nobody being able to check. Newman 2024 good

It is not subtle once you look. Of the 500-plus Americans on record as 110 or older, seven have a birth certificate. Worldwide, only about 18 percent of the most carefully validated supercentenarians have one, and in the United States that figure is zero. Newman 2024 A 2010 audit in Japan found that 82 percent of its 230,000 registered centenarians were dead or untraceable. NPR 2024 In Greece, an audit found at least 72 percent of centenarians were dead, missing, or collecting a pension they should not have been. One Japanese man feted as the country's oldest was found mummified in his bedroom, having died 32 years earlier while his family quietly cashed his pension checks. Newman essay

Even the beloved Blue Zones, the regions sold as longevity paradises, fit the pattern. Sardinia, Okinawa, and Ikaria turn out to rank below their own national averages on income, literacy, and life expectancy. Okinawa eats the fewest vegetables and has the highest obesity rate in Japan. And the single best predictor of where Okinawa's 100-year-olds live is which towns had their birth records destroyed by American bombing in the war. Newman essay good More bombing, more centenarians.

One fair caveat, kept in on purpose: not everyone agrees the rot is total. A geriatrician reviewing the work cautioned that it may overstate the error rate, and noted that some longevity research, such as studies of slow-aging families in the United States, does not lean on shaky birth records at all. NPR 2024 The point is not that no one is old. It is that the very oldest numbers, the ones used to argue about walls and limits, are the least trustworthy data in the whole field.

· · ·

Does aging ever stop?

Here is the genuinely unresolved science, and the honest answer is that the best datasets in the world disagree with each other.

Your risk of dying does not stay flat as you age. From your thirties onward it roughly doubles every eight years, a brutal exponential called the Gompertz curve. By 80 it is steep. The interesting question is what happens at the very top, past 105, where almost no one is left.

In 2018, a team led by demographers including James Vaupel and Kenneth Wachter tracked every Italian aged 105 and over for six years, nearly 4,000 carefully checked people. They found something strange: the death rate stopped climbing. It flattened into a plateau at roughly a coin-flip, about a 47 percent chance of dying each year, and it did not rise further with age. Barbi 2018 good A flat death rate is a profound thing, because a line that never climbs to certainty implies no fixed age where life must end.

If that plateau is real, it is the strongest argument against any fixed wall. But the case is genuinely shaky, and good scientists are on both sides.

· · ·

Is eating less the only hack?

Caloric restriction is the most-studied longevity idea there is. It is also the most oversold, and the part of this whole post where my own starting assumption took the biggest hit.

The dream is real in small animals. Feed a worm, a fly, or a mouse about a third less food and it can live dramatically longer. The trouble is that the effect shrinks as the animal gets bigger, and the two best tests in monkeys, our closest long-lived stand-ins, openly disagreed.

One trial, at the University of Wisconsin, found that 30 percent caloric restriction cut deaths from age-related disease roughly in half. Colman 2014 A nearly identical trial at the National Institute on Aging found no survival benefit at all; the restricted monkeys actually averaged about a year less. Mattison 2012 mixed When the teams pooled their data, the tie-breaker was almost comic: the Wisconsin "control" monkeys, the ones supposedly eating normally, were being fed a diet that was nearly 30 percent sugar. The restricted group partly just looked healthy because the comparison group was eating junk. Mattison 2017

So what about people? The one serious trial is CALERIE, which asked healthy, non-obese adults to cut calories by 25 percent for two years. The first finding was about willpower: even motivated volunteers with a research team behind them only managed about 12 percent, and it drifted back up over time. Ravussin 2015 The benefits they did get were real but ordinary: lower cholesterol, lower blood pressure, less inflammation, the same things you get from losing excess fat. Kraus 2019 good One epigenetic "aging clock" slowed by about 2 to 3 percent; two others did not move at all. Waziry 2023 mixed

And it came with a bill. The dieters lost measurable bone density at the hip and spine, and shed about 16 pounds, of which roughly five were muscle, not fat. Villareal 2016 good Most important of all: CALERIE never measured lifespan, and never could. The researchers say plainly that whether eating less makes a human live longer "will probably never be determined" by a trial. Ravussin 2015

The same goes for the pills sold as shortcuts to it. Rapamycin reliably extends life in mice but has no such proof in people. Metformin, the great hope, actually blunted the gains from exercise in one careful trial of older adults, dimming the very benefit people most want to keep. Konopka 2019 good Resveratrol and most "NAD boosters" are hype well ahead of evidence. The one genuinely strong drug result is mundane by comparison: the obesity drug semaglutide cut heart attacks and strokes by 20 percent in a large trial, but only in people who were overweight to begin with. Lincoff 2023 strong Once again, the win is really about reversing excess fat, not a secret of aging.

· · ·

What actually adds years

If eating less is oversold and the wall is unproven, what is left is almost disappointingly old-fashioned. The single most powerful lever is the one nobody markets as a hack.

It is fitness. In a study of 122,000 people that measured cardiorespiratory fitness on a treadmill and then followed who lived and who died, being in the least-fit group predicted death more strongly than smoking, diabetes, or heart disease did. Mandsager 2018 strong Read that again: in this data, being badly out of shape was a bigger killer than smoking. And unlike almost everything else, fitness showed no upper limit. The most fit kept gaining; there was no point where more became pointless.

So here is the whole field, ranked by how strong the human evidence actually is, with the hype stripped out.

• Be fit, and keep moving. Cardio and strength both, for as long as you can. Being unfit carried an even higher risk than smoking in the study above, and fitness has no point of diminishing returns. Mandsager 2018 strong
• Do not smoke. Just as essential and far better understood; quitting by 40 buys back almost a decade. strong
• Keep blood pressure and LDL cholesterol in check. Unglamorous, deadly when ignored, and well proven in trials. strong
• Do not carry excess fat. This is what most "diet hacks," fasting, and the obesity drugs are really doing. Lincoff 2023 strong
• Sleep, and stay connected to people. Loneliness carries a mortality risk on the order of smoking; both are easy to underrate. good
• Caloric restriction, fasting, rapamycin, metformin, NAD, resveratrol. Interesting, occasionally useful, wildly oversold. None has shown a longer human life. unproven in humans

Notice what the strong list has in common. None of it raises the ceiling. It works by not dying early, by clearing the heart attacks and cancers and accidents that take most people decades before any theoretical limit. That is the real shape of the thing. We are very good at helping more people approach the old ceiling, and we have no proven way to lift the ceiling itself.

The honest goal is not to find the cheat code for 120. It is to not get cheated out of 85.

Which leaves the wall exactly where we found it: unproven as biology, undeniable as a fact, and guarded by data too rotten to trust. Maybe someone breaks 122 this century; a serious forecast puts the odds above 99 percent by 2100. Pearce 2021 model Maybe the plateau is real and there is no wall at all. Either way, it has almost nothing to do with how long you will live. That part, unglamorously, is mostly up to the basics.

· · ·

Sources, and how I checked them

I started with two beliefs, a hard wall at 120 and eating less as the only hack. Both took damage. Before writing, every major claim was run through an adversarial fact-check whose job was to refute it, and the corrections were left visible above.