Counting the Elderly

March 1, 2012 3:33 PM

"Fewer persons alive at 70 today survive until 90 than forty years ago." True or False?

On the first page of the Introduction to the cookbook Nourishing Traditions (published in 2001) the authors claim, "Fewer persons alive at 70 today survive until 90 than forty years ago," but do not cite any sources or offer any explanation as to how they came by this statistic (at least as far as we've gotten on our reading of the book). Is this true? How would one support the claim or disprove it?

posted by jrb223 to science & nature (13 answers total)

The CDC measures life expectancy at different ages for US residents, and has for decades. This chart shows that the claim is false for US residents. You might be able to find similar data for international populations from their governments or from international health agencies.
posted by decathecting at 3:41 PM on March 1, 2012 [2 favorites]

That doesn't make any sense, and is easily disproved. More people survive to age 90 today than ever before in the US; all of those people were previously 70.

Now, if they mean "worldwide" I don't know how you could figure that, especially seeing how fragmentary records were in many of the most populous nations in 1961 (India and China and Brazil in particular).

If they mean "in the US" but they mean "among the set of people who were 70 in 1961, there was a higher percentage of people who made it to 90 in 1981 than there was of people who made it to 90 in 2001 who had been 70 in 1981" I am not sure how you would measure that, or how definitive it would be of anything except itself.

My guess is that they're on some kind of shtick about how terrible preservatives and processed food are, but they're a little out of date for the US, where processed foods have been the norm among people who could afford them since the nineteen-teens. (My own grandmother and great-aunts, born in the 1890s, ate canned vegetables, Wonder bread, and Hershey bars by preference and all lived until their late 90s.)
posted by Sidhedevil at 3:42 PM on March 1, 2012 [5 favorites]

sidhedevil: imagine that previously 20 percent of people survived to 70 and 4 percent made it to 90, and that now those numbers are 30 and 4.5. Then previously 20 percent of those who made it to 70 made it to 90, and now 15 percent do. Of course more people are making it to both ages, so like you I fail to see how this is a bad thing.
posted by madcaptenor at 3:50 PM on March 1, 2012 [1 favorite]

If they mean "in the US" but they mean "among the set of people who were 70 in 1961, there was a higher percentage of people who made it to 90 in 1981 than there was of people who made it to 90 in 2001 who had been 70 in 1981" I am not sure how you would measure that, or how definitive it would be of anything except itself.

That is what I understand the quotation/statement to mean.

It seems intuitively plausible to me. Back then you had to be more awesome to make it to 70, and now everyone and their grandmother hits 70, so the pool of people at that age is weaker -- and thus less likely to hit 90 -- now than it was before.

But it seems very unlikely to be true, given the chart on page 133 of decathecting's link.
posted by J. Wilson at 3:54 PM on March 1, 2012 [1 favorite]

I figured they had probably misstated what they meant, and that it was something along the lines of what you folks said, madcaptenor and J. Wilson.

But as decathecting's link points out, even if they meant that, they were inaccurate. And even if they had been accurate, it's hard to see what nutrition-related argument they were making would have been supported by that assertion.
posted by Sidhedevil at 4:04 PM on March 1, 2012

Just confirming what everybody else said, this is totally untrue. Coincidentally, I just finished a research paper on just this topic.

For example, in 1900, there were 3.1 million Americans over the age of 65. By 2008, this number had increased more than tenfold to 39 million, corresponding to an increase in the percentage of total population from 3.1% to 35%. The population over age 85 grew even more dramatically, from just over 100,000 in 1900 to 5.7 million in 2008 [2].

[1] Bell F., Miller M. Life Tables for the United States Social Security Area 1900-2100. 2005;
Available at: http://www.ssa.gov/oact/NOTES/as120/LifeTables_Body.html. Accessed 11/17,
2011.

[2] Older Americans 2010: Key Indicators of Well-Being. Washington, DC: Federal Interagency
Forum on Aging-Related Statistics. 2010.
posted by TungstenChef at 4:05 PM on March 1, 2012 [3 favorites]

I say false. In the fifties and sixties, if you had made it through all the childhood diseases and didn't get cancer, you would tend to be in the clear for your old age. But they can't be talking about fewer persons doing that today, because the same logic applies to today's population. They might be talking about a lower percentage of the population. And even if that were true, it would be not only because the general population is higher, but because most of those who die today between the ages of 70 and 89 would have been dead in their 50s or 60s in the old days. So it appears to say that people don't live as long as they used to, but we all know that the opposite is true.
posted by weapons-grade pandemonium at 4:13 PM on March 1, 2012

Read literally, the statement is nonsensical, for the reasons below.

"If they mean "in the US" but they mean "among the set of people who were 70 in 1961, there was a higher percentage of people who made it to 90 in 1981 than there was of people who made it to 90 in 2001 who had been 70 in 1981" I am not sure how you would measure that, or how definitive it would be of anything except itself."

What you're referring to would be a "cohort life table."

To understand the table posted by decathecting, you need to understand the difference between period life tables and cohort life tables. A period life table is a snapshot in time. You might not think so, because it appears to show how many more years a person aged 60 will live on average (implying that it is a statement about future mortality), but that is not the proper interpretation. The way to interpret it is the table shows how many more years a 60-yr-old will live assuming that death rates stay the same. So, a period life table for 2010 shows how long people would live if everybody were to live their entire lives under 2010 death rates.

But of course, death rates change over time, so a period life table is strictly hypothetical. It is a snapshot of the death rates that exist frozen at that point in time.

By contrast, a cohort life table is what you might think a life table is: It is an actual tabulation of how long a group of people born in a given year have lived. So, a 1900 cohort life table will show you how long, on average, people born in 1900 lived, and how much longer people who were aged 60 in 1960 lived, and so forth. Note that it is only possible to construct a complete cohort life table (that is, a table going to the oldest ages) for people who were born a long time ago.

Not surprisingly, the vast majority of life tables you see -- like the one decathecting posted -- are period life tables. Accordingly, it doesn't actually address the issue raised in the quote above. For that, you want to look at a cohort life table.

Returning to the statement in the OP, if you read it literally, it's nonsensical. It refers to "persons alive at 70 today" and makes a statement about how many survive to 90. We don't know how many will survive to 90, because they're only 70 today! At best, we can either (1) make a statement about how long they would live if death rates did not change from today's; or (2) make some kind of forecasts about how death rates will change in the next 20 years.

If it is (1), it is false, for the US anyway. Death rates have declined at all ages. Under today's death rates, more 60-yr-olds would survive to 90 than would have under the death rates forty years ago.

And if it is (2), it is most likely false. The gold standard for mortality forecasts (Lee-Carter), as well as almost all other forecasts, predict that death rates at all ages will continue to decline for the foreseeable future.

(PS - I have a PhD in Demography specializing in mortality, for what it's worth; Lee was my dissertation advisor.)
posted by mikeand1 at 4:34 PM on March 1, 2012 [21 favorites]

Adding: It is almost certainly false worldwide as well.

Death rates at all ages have declined over the last forty years in almost every country that keeps statistics. That includes China and India, who hold a huge chunk of the world's population. There may be some anomalies in age-specific death rates here and there, but the overall trend is abundantly clear. (The anomaly for China would be the so-called "Great Leap Forward" in the early 60s, during which mortality soared, but even the effects from that have long since subsided.)
posted by mikeand1 at 4:39 PM on March 1, 2012

Based on current data this claim is nonsense, as has already been pretty thoroughly debuncked, but that isn't to say that it might not end up being accidentally true in the end. A particularly virulent and transmissible influenza virus during peak season, an unpredictable zoonotic pandemic, or other game changing infectious disease could easily shift life expectancies dramatically.

Most of our elder-care infrastructure is dramatically unprepared for 100 year type infectious disease events, and is far more centralized than it has ever been in human history.
posted by Blasdelb at 6:26 PM on March 1, 2012

I think what they're saying (in the most slanted way possible) is that the number of people who live to be seventy has increased faster than the number of people who live to be ninety.

Either they're very bad at calculus; are prescribing a diet that will kill you in your 50's and 60's unless you are in spectacular health of just making up statistics.
posted by Kid Charlemagne at 7:15 PM on March 1, 2012 [1 favorite]

It seems to me that the claim is trivially true. Some people who were 70 years old 40 years ago went on to survive to age 90, but no one who is 70 years old right now has done so.
posted by baf at 7:43 PM on March 1, 2012 [1 favorite]

It's at least plausible.

The people who made it to 70 in 1962 were probably, on average, of genetically healthier stock than the people who make it to 70 today. So the share of those people who went on to live another 20 years might have been higher.

Get to 70 in 1962 and you're probably not obese, diabetic or particularly prone to heart disease. You're probably going to die of one of the diseases that strike old people disproportionately (cancer, etc).

Make it to 70 today and you may well be diabetic or have already had a couple bypass surgeries. We're pretty good at managing those things today, so *lots* more people make it to 70. Ultimately those things tend to catch up with you well before you make it to 90, though.

Easy enough to disprove with the data mentioned above, though.
posted by pjaust at 7:58 AM on March 2, 2012