Jeanne Calment - Fraud, lived only till 99 years of age

[TomBAvoider]

It appears a case has been made that JC took her mother's identity in order to avoid inheritance tax, and that is how she could claim such an old age - in reality, she hit 99. That would make the oldest human the American woman, Knauss at 119:

https://www.thetimes.co.uk/article/woman-lived-as-122-year-old-mother-to-dodge-tax-cfwd8bcm6

Of course, the case is by no means any kind of proof - it may well be that the debunking is no good, and in fact JC was legit. 

[Clinton]

I read about the controversy of JC’s age years ago and thought her nearly 123 year age was suspicious.

When you consider the lives of multiple billion people (N > 10 billion) and the difference in lifespan between oldest and second oldest is approximately 3 years or 2.5%  it immediately looks statistically speaking as nonesense;  even 2.5 months difference between lifespan of top 2 individuals of that size of population is enormous and strange.

[mccoy]

On 1/2/2019 at 5:13 PM, Clinton said:

I read about the controversy of JC’s age years ago and thought her nearly 123 year age was suspicious.

When you consider the lives of multiple billion people (N > 10 billion) and the difference in lifespan between oldest and second oldest is approximately 3 years or 2.5%  it immediately looks statistically speaking as nonesense;  even 2.5 months difference between lifespan of top 2 individuals of that size of population is enormous and strange.

Clinton, I don't know if you are an expert in statistics, but statistically speaking JC would make sense as an outlier, the most conspicuos outlier, as a matter of fact, the 100% percentile, the absolute upper bound of the population age interval so far. Actuarial science is interesting in that it deals with the whole population, not just samples.

Of course a statistical analysis she wouldn't be considered, unless maybe dealing with a distribution of extremes.

It's still to be ascertained if the fraud claim is legit

 

[Clinton]

It’s been about 20 years since I studied statistics ... I may be off with my expectation.  Even #2 (Sarah Knauss) was an outlier; everyone after her and JC died at 117 or earlier.

But beyond SK and JC the lifespans are all within a few weeks of each other.  With N = many billion I expected JC and SK lifespan to be all closer together to together like what we see from the 3rd longest lifespan and down.

Statistics is something I’d like to re-study to assist me in skimming over scientific journals, etc..

[mccoy]

I understand your reasoning, but maybe there are quite a few people who woudl fill the gaps, people unkown to the actuarians because they were born in the Congo or in the deep jungle of Bolivia or in the western world but have no official birth certificates anyway.

[Saul]

Yeah -- zillions of them.

[TomBAvoider]

Even if there are a lot of people "deep in the jungle of Bolivia", born in "Congo" etc., I think we are on pretty solid ground with the data we already have. I find it unconvincing that it just so happened that *all* the lifespans that fill the gap between 117 and 122 years happen to be the unaccounted from jungles of Bolivia and Congo. You would imagine that if the distribution is equal, then at least a substantial portion of those would fall into the hundreds of millions of records we actually do have. A sample size that we have - of hundreds of millions of records - is big enough. I don't think that doubling that sample size would give us different results - same as that a hundred million records never showing a human taller than 10 feet is enough - making the sample 10 billion will not suddenly turn up a guy/gal that's 10 feel tall. 

We have enough records to draw conclusions about natural livespan of the very old - and adding the unaccounted for in the jugles of Bolivia or Congo would not add to the data any more than it would turn up a 10 foot tall human. Plus, frankly, I suspect that if anything, odds of a superlong lived human in the jungles of Bolivia or Congo are lower than in more complex societies, in the same way that extreme longevity among mice are more to be expected in the lab than out there in the wild. 

Bottom line, for at least 3 very good reasons, I suspect that adding info from the jugles of Bolivia would not change a single thing, statistically. YMMV.

[mccoy]

The issue of Bolivia and Congo was raised to hypothesize tha lack of a more continuos distribution, with less spacing between extreme data. However, outliers being what they are, some extreme data distanced between themselves is probably nothing strange.

[mccoy]

I found there is a thriving branch of statistics which tries to infer the maximum age attainable by populationsby present data

For US females it has been estimated to lie in the range 126-129 years

Above is illustrated the n=309 sample, with the maximum value of 119 years. We can easily see that the highest 5 data are pretty distanced between themselves, with the top 2 pretty distant. 115 years presently seems to be the highest age beyond which occurrances are extremely rare, according to this US database.

 

Who wants to live forever? An analysis of the maximum lifespan in the US

 

[TomBAvoider]

5 hours ago, mccoy said:

The issue of Bolivia and Congo was raised to hypothesize tha lack of a more continuos distribution, with less spacing between extreme data. However, outliers being what they are, some extreme data distanced between themselves is probably nothing strange.

But that still doesn't answer my objection. If you say that Bolivia and Congo jungles have all the people between ages of 119+ and 122- I ask - why are those 3 years ONLY found in Bolivia and Congo jungles? If the distribution of lifespan is random across the world, there is no reason why SOME of the *between* 119-122 group would not be also in the records keeping areas... why are ALL the 119-122 ONLY in Bolivia and Congo jungles - It Does Not Make Sense. Given random distribution, we should see SOME 119-122 in Bolivia and Congo, and SOME in Europe/America/Japan etc. Sorry, but statistically this is nonsense. It's as strange as if you told me that people who died at 25, 26, 27, ONLY lived in Congo and Bolivia, and no person ever died at 25, 26, 27 in Europe/America/Japan and people in Europe/America/Japan died in every other age, but never at 25, 26, 27, and for those deaths you have to go to Congo and Bolivia jungles. It is 100% nonsense. 

The other explanations for why the between 119+ to 122- group only happen in Boivia and Congo would be sample size. But that doesn't work - because the sample size of hundreds of millions across the world is a *statistically* valid sample. You don't need more. Same as the example I gave before - to establish the range of maximum height of a human, it is enough to sample a subsection of humans - by getting to sample more people, you will not suddenly find a 100 meter tall person. Getting more people hidden in jungles is not going to turn up a 100 meter man. That is the whole point of statistical *samples* - otherwise you would never even have such a concept.

Bottom line, having such a large gap between 119 and 122 is definitely fishy. If max lifespans are tightly represented at the very extreme, then having such a whopping jump is certainly eyebrow-raising because at the very tip of that extreme, every month makes a difference, so to jump by THREE WHOLE YEARS and that is the ONLY example, is concerning. Somebody has to be the oldest - if even by a fraction of a second. But we grow worried when that oldest is older by a giant leap that is never even come close to by any other human ever reliably recorded.

I therefore am not surprised that someone is looking into the Calment case. It's the old principle: extraordinary claims require extraordinary proof. The question is - have we been supplied with extraordinary proof in the case of Jeanne Calment? Inquiring minds want to know 🙂

[mccoy]

TomB, I would not take literally Bolivia and the Congo, they were just examples of remote countries, whose life conditions, as you rightly point out, are not even too favourable to longevity..

Statistically speaking, a jump of 3 years does not appear to consitute an anomaly. If you read the previously quoted article, statistical inference by the extreme values model illustrated in the article would show that just in America maximum age of 126 to 129 ages would be expected (mathematical inference from the n=309 existing sample of ultracentenarians), even higher than Jean Calment's age.

[TomBAvoider]

And my point would still stand - divide the world between "recorded lifespans" and "unrecorded lifespans" - Bolivia, Congo and everything else would be in "unrecorded lifespans". My argument still stands - why is it that ages between 119 - 122 supposedly ONLY occur in the "unrecorded lifespans"? There is no explanation. The speculation about theoretical max lifespan in the U.S. is unfortunately hampered by a rather awkward fact - there's nobody that exceeded 119 in the U.S. Even if we were to accept theoretical max as 126/129 I don't see a mathematical justification for lifespan distribution that's anything but even - big jumps based on what? I don't think so. Btw. it's not encouraging that they accept the case of Jeanne Calment uncritically - it starts looking like begging the question... hey, if they really want to get the numbers up, they should accept the various tall tales from the Caucas region where regular claims are made of people hitting 130 and more (when people visit the same supposedly old folks, those folks claim to be 10 years older just a year later... funny business).

[Todd Allen]

I agree with Tom, the big gap seems suspicious.  Even if Jeanne Calment's age is legit I'd say as a freakish outlier there is minimal relevance to current human lifespan.

[mccoy]

19 hours ago, Todd Allen said:

I agree with Tom, the big gap seems suspicious.  Even if Jeanne Calment's age is legit I'd say as a freakish outlier there is minimal relevance to current human lifespan.

That's another line of reasoning. For example, based on the posted database of n=309 American females, maximum expected lifespan would be 115 years, all extremely rare values above that being considered as total outliers. this doesn't imply that such values are not genuine, simply that they are so distanced from the others (one definition of outliers) that  they are caused by an incredibly rare, almost impossible set of combinations. 

[TomBAvoider]

Al just posted this on his thread:

 

Evidence that Jeanne Calment died in 1934, not 1997.
Zak N.
Rejuvenation Res. 2019 Jan 30. doi: 10.1089/rej.2018.2167. [Epub ahead of print]
PMID: 30696353
Abstract
I present a body of data that, I argue, cumulatively casts serious doubt on the validity of Jeanne Calment's accepted world record of human lifespan. First, I assess the plausibility of the record based on the lifespans of other centenarians in the International Database of Longevity (IDL) and critique some arguments put forward previously in support of this record, including the longevity of Calment's ancestors. Second, I review the literature dedicated to Calment and discuss multiple contradictions in her interviews, biographies, photos, and documents. I argue that the evidence from these sources motivates renewed consideration of the previously rejected hypothesis that Jeanne's daughter Yvonne acquired her mother's identity after her death in order to avoid paying inheritance tax and that Jeanne Calment's death was reported as Yvonne's death in 1934. Finally I discuss the importance of reconsidering the principles of validation, due to the possibility of similar problems regarding other exceptionally long-lived people and the mistaken inferences that researchers may draw from flawed datasets. The phenomenon of Jeanne Calment may prove to be an instructive example of the uncertainty of seemingly well-established facts.

[mccoy]

If the doubts will be proven to be true, then Sarah Knauss is going to hold the record at 119 years.