The Myth of Japanese Longevity | Hacker News Reader

The Myth of Japanese Longevity

asserttrue.blogspot.com

144 pointstechdog65 comments

Necessary context for this article can be found in an earlier Hacker News comment

http://news.ycombinator.com/item?id=5196734

by another participant on an earlier article from the same blog:

> > It appears you've made some sort of resolution to publish and promote a blog entry per day in 2013. 40 entries in 41 days this year vs. 46 in all of 2012. You should reconsider - whatever your reasons were, I doubt they included a desire to develop a reputation for presenting topics that were sensationalized and thinly researched [1] produced with a pace that ensures discredited theories dont get reviewed.

[1] http://asserttrue.blogspot.com/2013/02/drug-companies-stop-h...

> Wow, nice spot and they have all been submitted to HN. I have never seen anyone's submission history be so hell bent on self promotion:

https://news.ycombinator.com/submitted?id=techdog

That was followed up by another set of comments:

http://news.ycombinator.com/item?id=5240084

> I'm beginning to flag these posts.

I agree.

What the hell do placebos have to do with this article?

And I still think the most important part of that article was that placebo ingredients were not included in studies, making them hard to understand/reproduce. Wasn't that completely accurate? Set aside any implications that may or may not have been there.

I'll give that it's probably poorly researched, but what you're linking is only half-relevant to that.

Good links. Looking at past posts some are a little out there (Read crazy conspiracy theories)

Shame, it's an interesting idea. Could we be mistaken and the Japanese don't live longer. I believe similar mistakes are happening in blue zone theory.

But I'm a firm believer that some people can just can make a good argument about anything.

So if they have been very wrong in the past, or very wrong in part of their argument then I take it as now they are wrong and just making a good argument that's fooling my flawed brain.

Oh noes, conspiracy-related posts. Closing the tabs to his site right away... or not.

Randomly plugging numbers. Say there are like 250,000 'undead' people. Say that on average they've been on dead for average of 50 years (crazy! I know!). This works out to an extra 12.5 million years of life. There are 127 million Japanese. If we assume average life expectancy of like 70 years, works out to 8.9 billion years of life. If we drop those 12.5 million or so years that don't exist... we're still at like 8.9 billion years.

Well to plug more numbers, those 250,000 'undead' people having been 'undead' for the last 20 years collecting pension checks amounting to say, $20k/year would mean:

    5 million 'years' * $20k/year = $100 billion

Japan's pension fund is severely underfunded, so even this amount would help.

It's true that the 'undead' wouldn't skew the mean life span much. But given the way life expectancy is calculated (http://en.wikipedia.org/wiki/Life_expectancy), it ends up being closer to median life span - so an overstated number of centenarians (people over 100) could significantly skew things.

And it's worth noting that while other developed nations are close to Japan in terms of life expectancy, they're WAY behind in terms of number of centenarians. Switzerland for example has an average life expectancy of 81.8 years, vs. Japan's 82.7. But Japan supposedly has 3.5x as many centenarians per capita (http://en.wikipedia.org/wiki/Centenarian#Centenarian_populat...).

> it ends up being closer to median life span - so an overstated number of centenarians (people over 100) could significantly skew things.

Could you elaborate? Isn't the median always _less_ sensitive to outliers compared to the mean?

Not always, no. They measure different things and fail in different ways.

Take the dataset [7, 8, 9], with a median and mean of 8. Adding a 100 to the set results in a median of 8.5 and a mean of 31, so the mean moves much farther. This is probably the effect you're thinking of: the mean can take extreme values into account "too much".

But I can also make the median move more. Take the dataset [0, 50, 100]. The median and mean are both 50. If I add [100, 100] to it so it becomes [0, 50, 100, 100, 100], the mean moves to only 70, but the median moves all the way up to 100! There was a "gap" in the numerical sequence that the median could jump over, but the mean couldn't.

Here's a different way of moving the median further. Take the dataset [1, 1, 1, 1, 2, 3, 4, 5, 5, 5, 5]. Bathtub-shaped data. As I add fives to the set, the mean goes 3, 3.17, 3.3. But the median goes 3, 3.5, 4! Medians move past thin spots in distributions very quickly.

Mean is sensitive to distant outliers; median is sensitive to unevenly distributed data and numerical gaps.

To come back on topic, while I don't have a reference for the age-at-death distribution, I think it's bathtub-shaped. Hence, the median might be more sensitive to extra values at the top than the mean would be.

While those are interesting counterexamples, they don't come close to modeling a realistic "age-at-death" distribution. There isn't any such distribution whose median would be significantly skewed by a tiny minority of centenarian outliers. These distributions are basically unimodal with the exception of a some degree of infant mortality, and the average (median or mean) is not located at a thin spot in the distribution (quite the opposite) [1].

[1] http://www.longevitas.co.uk/site/informationmatrix/mortality...

Ah! Thank you! I had the most miserable time trying to google for that exact graph.

No, it's not "closer to median life span."

Expected value is a mean.

In the case of life expectancy at age X, it is the mean, conditional on having already survived to age X. (In other words, life expectancy at age 20 excludes everyone who died between birth and age 20.)

This is unbelievable: it looks like they are faking child births, too! The rate is up a whopping 15%.

http://web.archive.org/web/20120203092942/https://www.cia.go...

https://www.cia.gov/library/publications/the-world-factbook/...

Also, from my experience smoking seems to be rare both in Tokyo and the countryside, so I'd like the see some statistics about Japanese being "smoking fiends" before I assertTrue().

A fine place to start of course:

http://en.wikipedia.org/wiki/Smoking_in_Japan

25% of the population smokes, and their rate among adults is roughly 55% higher than in the US.

http://en.wikipedia.org/wiki/Smoking_in_Japan#Prevalence

Males: 37%, females: 9%.

Compare with 22/17, M/F in the USA.

[edit] Beaten to the punch, apologies for the double post.

Rare in Tokyo? Are you kidding? They even have public smoking booths so that people can smoke in otherwise restricted areas.

He was maybe confused since it's now forbidden to smoke in many large public areas outside of confined spaces marked off for smoking, like the areas around train stations. There's less smoking-and-walking in general.

An exception to this that I wasn't aware of until just last month is the boarding platform for the bullet trains in some stations. I had to walk through a dense cloud of cigarette smoke on my way to my bullet train car at 9am on a Sunday. That was not pleasant.

From my understanding and observations when I visited Japan for a couple of weeks, Japanese people don't eat/drink/smoke while walking. Even for outside vendors, they will eat it on a nearby bench or stand and eat it before moving again.

I did see smokers often in the smoking areas. More than I see in NYC.

>Japanese people don't eat/drink/smoke while walking.

What about all those Japanese schoolgirls that run to school with a piece of toast hanging out of their mouths because they overslept?

I can't find the data for faking child births from the links, but since the Democratic Party instituted "Child Credits" between 2009 and 2012 (iirc), where families would receive something like $200/child/month, I wouldn't be surprised if people were faking births.

This article is sensationalist and false. If one reads the CBS page linked from the article (http://www.cbsnews.com/stories/2010/09/10/world/main6853038....), we find the following:

"An official at the Health Ministry's statistics bureau said Friday's survey does not change Japan's status as a fast-aging nation because life expectancy calculations are not based on family registration records."

If it isn't based on family registration records then what is it based on?

If there's one thing people know about the Japanese, it is how much they like to save face.

Methinks that the page has been defaced. This is what I saw when I opened the link: http://i.imgur.com/WHHxE7n.png NSFW

Viewing the page source shows the following as the background image: http://www.blogblog.com/1kt/watermark/body_background_birds.... NSFW

Interestingly, the same page, viewed in xombrero, shows nothing amiss.

Yet in Chromium, the following is what I see when I load the background image as linked to in the page source: http://i.imgur.com/kJ5veiA.png NSFW

I'm not the only one who sees it.

Methinks it's just you. When I visit the page, I see nothing of the kind. No one else seems to have noticed the butt background either.

I have received verification from someone in Great Britain, that they also see the butt background.

Are you on an usecured wifi network or something?

Because I see no such background. I feel like maybe you are being MITMed.

My connection is as secure as T-Mobile makes their "4G Mobile Hotspot."

The fact that an acquaintance in England (I'm in Hawaii) also saw the x-rated background leads me to believe that something else is going on with this page.

Weak article. They assert there is a problem with the data, point out at least one flaw, conclude that the entire data set is wrong and then attempt to draw conclusions from it. The author seems to be trying to use negative evidence, i.e. induction, to prove something, like in the following example:

Argument: The average age of a population is 10.

Counter-argument: At least on one of the numbers is greater than it should be.

Conclusion: The average age of a population is 8.

The meat of the article is:

>the fact that there are 234,000 unrecorded deaths in the Japanese population means the often-touted life-expectancy figure of 82 years for Japan now has to be considered suspect

Not sure how you can argue against that reasoning unless you can show Japan's life expectancy calculations have nothing to do with official estimates of how many people are currently older than 100 (which is possible).

Your example doesn't follow the spirit of their argument at all. They aren't proposing their own average, and their evidence involves far more than one data point.

The headline of the article is proposing an average, though. By saying "the myth of x" they insist that x is false. In this case, this is equivalent to asserting that Japanese life expectancy is not above the average in Western nations. Seems like simply a case of a misleading or exaggerated headline.

I think your conclusion should be re-stated as

Conclusion: The average age of a population is less than 10.

Which would be true if the counter-argument was found to be correct, would it not?

This is about Japan in general as well. Specifically it's in Okinawa where people supposedly do live an unusually long time, with the rest of the country more unspectacular. I would like to see whether this is true in Okinawa as well.

" Sogen Kato, Tokyo's oldest man, as found on his 111th birthday. Kato did not respond to our requests for an interview. "

hidden joke. I'm glad I caught it.

There's something fishy going on with the page. http://news.ycombinator.com/item?id=5326345

No this is not entirely true. The Kyoto region of Japan, was a subject of survey sometime back, I can assure you that I read it myself, and it included few more regions of the world proven by a UN collaborated study.

That study is mentioned here: http://en.wikipedia.org/wiki/Supercentenarian

also, see the verified records http://en.wikipedia.org/wiki/Oldest_people

A extended list: http://www.grg.org/Adams/E.HTM

Publications: http://www.supercentenarians.org/publications.htm

"CIA's web page on Japan's death rate shows Japanese mortality as having dropped by 10% in one year, in 2012"

10% sounds like a lot to me, but that 2011 earthquake/tsunami will have caused a peak in death rate. Likely, there also was a stress related peak outside the directly affected areas. That peak would be followed by a through, just like one sees elsewhere after a hot summer or harsh winter (http://en.wikipedia.org/wiki/Mortality_displacement)

Ten percent fewer deaths would be, I guess, about 60,000 'excess' deaths in the year of the tsunami and 60,000 fewer in the next year. Does anybody know whether that would be feasible?

The number of deaths from the tsunami is listed at about 16k (http://en.wikipedia.org/wiki/2011_T%C5%8Dhoku_earthquake_and...). I'm surprised it was so high, but in general deaths from natural disasters are low in number, compared to regular things people die from like disease in old age.

That is the number of direct victims. My point is that there will have been people (elderly or otherwise physically weaker) who died a few weeks/months earlier than they would have if that earthquake hadn't struck, for example because they got a heart attack worrying whether their grandchildren survived.

Such deaths both increase mortality in 2011 and decrease it in some time later.

While there may have been around 25000 deaths, what I think happened is that there were many records destroyed and many people displaced, so there might have been much discrepancy in the collecting of accurate data.

What's interesting to me isn't the fraud but the fact that the fraudsters keep the dead people around. Why not bury them in secret?

They are more likely living in somewhere relatively dense in population, and it is easy for people to take notice something weird going on with noise, and more often, report of odors afterward. (Unless they go out somewhere in the middle of the mountain, which happens; not everyone having a car can make this challenging option.)

Another factor is their mindset, too. While most of Japanese are not religious, they do follow certain cultural customs from Buddism and Shintoism, etc.

> report of odors afterward

... which is one of the big reasons I'd expect them to get rid of the bodies instead of keeping them around.

Actually, I've heard a many cases of Japanese news that reveals the existence of the body in the ground because someone buried the body in their yard or in the ground under their house, etc. I'd still think it'd be challenging (and probably very unpleasing) to do the same while they are in possession, but at least they can have some effort of it contained.

That's hardly new, the Japanese have known for quite some time that their records about the elderly are "a little bit" off.

It's new to western media though. There are countless articles floating around about Japanese longevity, attempting to connect it to diet and exercise, oblivious of this pension fraud factor:

http://www.guardian.co.uk/world/2011/aug/30/japan-life-expec...

http://abcnews.go.com/Health/key-longevity-chatting-japans-c...

http://observer.com/2010/08/revealed-the-secret-of-japanese-...

http://www.lef.org/magazine/mag2011/may2011_The-Little-Known...

And there have even been funded studies and papers published in medical journals about this!: http://www.ncbi.nlm.nih.gov/pubmed/18924533

Just because there's welfare fraud in Japan doesn't mean their life expectancy can't still be higher than in other countries. Also, let's not pretend there's no welfare fraud in other countries.

Tax fraud too. I rented a house in 2002 from a man who died in 1988.

"It's new to western media though."

I'd be surprised, because western media is exactly from where I found out about this two years ago or so:

http://www.nytimes.com/2010/08/15/world/asia/15japan.html?_r...

(...and a few others.)

Maybe not news to you, but in american media the "Super healthy asian food" image sells a lot of snake oil. In fact, there are people who think going to a restaurant/mall and getting asian food counts as a diet and/or eating healthy. They have no idea how different "westernized" asian food is to how it is(was?) made in their countries of origin - or that orange chicken, sweet & sour pork and chicken-fried rice might not be the best choice for eating healthy/dieting.

While this article makes it somewhat sensationalistic linking it to welfare fraud, perhaps, the majority of "unregistered deaths" are the ones dying alone.

There are so many cases in Japan that people die alone, and people not noticing it until their neighbors reports unusual odors, which may or may not happen. (although census should supposed to be catching that...)

Our medicare/medicaid scam artists have a thing or two to learn.

From the linked article in the OP:

Move on, nothing to see here.

20/~100 of the verified oldest people were japanese according to

http://en.wikipedia.org/wiki/List_of_the_verified_oldest_peo...

That's a handsome statistic right there.

Interesting. Sounds almost like a conspiracy though, it would be nice to see some confirmation. If it's true, then a lot of government numbers will need to be updated, and it could have impact on health studies.

It's known for some years: http://www.nytimes.com/2010/08/15/world/asia/15japan.html?_r....

on the subject of conspiracy, suicide rates in Japan are widely rumoured to be gamed by the authorities to be essentially "fixed" at a little north of 30,000/yr and many suspect that a proper classification of "missing" or "suspicious deaths" to suicides would increase this number to 100,000/yr.

It is bunk that Japan is known for good bookkeeping. Some large Japanese financials are still using paper-based risk assessment and transactions. Some are losing millions of dollars each month to human error.

Is there some sort of campaign against the Japanese healthcare system now?

Wow!! Shocked!!!

Maybe, maybe not.

The top cause of death in Japan is Amyloidosis:

http://www.nationmaster.com/country/ja-japan/mor-mortality

Amyloidosis, in case you're wondering, is a protein disorder most commonly found in super-centarians (people 110 years old or older):

http://en.wikipedia.org/wiki/Amyloidosis

So not only are the welfare fraudsters great at hacking the social support system without getting caught, they're also really knowledgeable about arcane disease patterns in super-centarians, just to give their lie that extra ring of truth once they decide to finally leave the dole. And... somehow... they're conning medical professionals into reporting this as a cause of death for mummified elders.

While a large number on the face of it, I'm not even convinced 230,000 missing elderly is statistically significant given the size of the population. [1]

I usually give articles a stronger benefit of the doubt, but this is from the site that's arguing that lung cancer isn't really related to smoking, and questionable claims about the causes of autism.[2]

Now I'm wondering, is the blog's title an easter egg? Is the whole point to just take some crazy proposition and see how many people will buy it? Is this entire blog just trolling the internet?

[1] EDIT: Someone ran the numbers, thanks icegreentea http://news.ycombinator.com/item?id=5326622

[2] Hacker News discussions raised interesting counterpoints, questioning large gaps reasoning in previous pieces: http://news.ycombinator.com/item?id=5196579

Perhaps the best comment: http://news.ycombinator.com/item?id=5196734

"It appears you've made some sort of resolution to publish and promote a blog entry per day in 2013. 40 entries in 41 days this year vs. 46 in all of 2012. You should reconsider - whatever your reasons were, I doubt they included a desire to develop a reputation for presenting topics that were sensationalized and thinly researched produced with a pace that ensures discredited theories dont get reviewed."

Umm, look more closely at your link. Amyloidosis is only "top" because the list is alphabetical.

How the hell does someone die from "Flatulence and related conditions"?

So... what I said about it being "top" was technically correct, no?

Jk, good catch, monochromatic.

I'd like to say I wouldn't be making such sloppy errors if assertTrue() hasn't been engaged in this series of nutty claims recently, but that's probably giving myself too much credit.

Thanks for keeping me honest.

Reply on news.ycombinator.com