But there are loads of people who hit 100. The random google search talks about 450k centenarians.
100 years ago is 1922. Perhaps the 100 year old person in the 70s is hard to believe, but 1922 is well into major bureaucracy existing, and where it is harder to just make up a person. There are of course loads of stories of identity fraud and the like, but by the time you would want to execute the fraud (I guess in your 40s?) you're looking at trying to do ID fraud in the 1960s.
For the super extreme cases, it maybe feels worth trying to pull this trick off. But at one point we have confirmed that getting to 100 is a thing.
I didn't literally mean that people don't ever make it to 100, just that it seems plausible that incorrect reports (fraud, error, etc.) are orders of magnitude more common on a global level. Imagine trying to study longevity with data that has 4M or 40M erroneous centenarians and only 450k real ones.
Some of these isolated 'blue zones' are cults/small communities that may also have a conflict of interest in convincing people that their way of life leads to longevity.
Compare to the total number of records, very, very few were destroyed in WW2. It's not like every building on the planet were burned down.
And lots of people have their own records in family bibles/other books. Modern genealogy sites (and even general web searches) have (and still are) digitizing old records, news papers, graveyards, and lots of other historical records, and this data is being fed into all sorts of databases/AI/analysis tools, making really precise records of giant family trees possible, with original photos, handwritten notes, census data, and on and on.
So there's likely a better, more accurate records of all of this, back hundreds of years, aggregated than ever before.
I know in my case, it doesn't take long now to trace parts of my family back 400-500 years into ancient France (I'm in the US). The number of original records I can see is astounding. It's only getting better.
And it's pretty easy statistically to count % of people making age 100 in countries or regions where records are very, very good, get a good model of how that works, including as many relevant variables as people want to study, and use that to get a good estimate of how many are in places where records are poorer quality. That is how science is routinely done.
100 years ago is 1922. Perhaps the 100 year old person in the 70s is hard to believe, but 1922 is well into major bureaucracy existing, and where it is harder to just make up a person. There are of course loads of stories of identity fraud and the like, but by the time you would want to execute the fraud (I guess in your 40s?) you're looking at trying to do ID fraud in the 1960s.
For the super extreme cases, it maybe feels worth trying to pull this trick off. But at one point we have confirmed that getting to 100 is a thing.