collision detection: Study: After 70, being better-educated means worse memory-loss

January 09, 2007

Study: After 70, being better-educated means worse memory-loss

Check this out: A new study found that people with high levels of education tend to have bigger declines in their "working memory" after age 70 than the less-educated.

To figure this out, a team of scientists took a whackload of data that a group called AHEAD has collected since 1923. Essentially, AHEAD gave word-memory tests to old people. They'd read 10 common nouns to the subjects, ask them how many they could remember, then ask them again five minutes later. When the scientists crunched the AHEAD data, what did they find? Above age 70, people who had high levels of education experienced much larger decreases in their ability to perform this task than people who weren't as well educated.

This inverts the well-worn concept that education always prevents your brain from going soft. Up until now, most studies have found that people with higher levels of education stay smarter as they age: Their cognitive skills remain sharper, and they experience less dementia. But apparently this may not hold true for their working memory -- their ability to temporarily store and manipulate information. This new group of scientists suspects that as the well-educated age, they're able to use their schooling to help compensate for normal, age-related memory deficiencies -- but this strategy somehow falls apart at age 70. As Dawn Alley, the lead author on the study, noted in a press release:

"Even though we find in this research that those with higher education do better on mental status tests that look for dementia-like symptoms, education does not protect against more normal, age-related declines, like those seen on memory tests," said lead author Dawn Alley of the University of Pennsylvania, who conducted the research while a doctoral student at the USC Davis School.

I'd like to know more about how the scientists classify "high levels of education", but unfortunately the article is behind a paywall.

Either way, it's an interesting gloss on the current rage for "brain training" tools for the elderly -- the idea that playing Sudoku, or Nintendo's Brain Age, can help keep your brain young even as you sail off into your golden years.

Posted by Clive Thompson at January 09, 2007 11:56 AM

Trackback Pings

TrackBack URL for this entry: http://www.collisiondetection.net/mt3/mt-tb.cgi/1608

Comments

I work in brain injury research, and we classify level of education using the follow options:

Grade 6

Grade 7

...etc...

Grade 12, no diploma

High School Diploma (or GED)

Work toward Bachelor's Degree

Bachelor's Degree

Work toward Associate's Degree

Associate's Degree

Work toward Master's

Master's Degree

Work toward Doctorate

Doctorate

Other

For the purposes of study recruitment, it is common to match control and experimental participants in these categories. Often the groupings are fuzzier than the specific list used here. Level of education often forms one of the major pillars of a study's demographic control, which is why, when reading a study that examines lifestyle factors or developmental issues, I'm always suspicious of the phrase, "Controlled for all other variables." :-)

Posted by: Jonathan Dobres at January 9, 2007 6:19 PM

Hey Clive -

I gotta admit. I wouldn't remember 10 nouns after 5 minutes, and I'm not even 40. Ask my wife - when I go shopping, I always get all the things she told me we *have*. And it gets worse when I'm reading a good book (or one of your blog posts), or even worse when I'm writing something, which just shows the negative effect of education. Then I'll agree to anything you ask me and 2 minutes later deny I ever heard you ask.

In short, could it be that educated people have more interesting things on their minds then 10 random nouns? :)

I wonder if anyone ever did the correlation between level of education and success in trivia shows...

cheers,

- Yishay

Posted by: yish at January 9, 2007 7:56 PM

(of course, they would have to control for brain scientists

Posted by: yish at January 9, 2007 7:57 PM

I glanced at the paper (yay for university web proxies) and the researchers plot their data using "years of school completed" as their measure of education level, with 4, 8, 12 and 16 years completed plotted. Since the sample starts with people at age 70, this is before the time of widespread kindergarten, so it means the bins are set for 4th, 8th, and 12th grade, and college graduates.

Also, the graph showing the largest absolute decline for most highly educated people, which is for verbal recall, seems to show almost identical relative decline normalized by score at 70 years old. For example, people educated 16 years declined from a score of 4.6 (whatever that means) at age 70 to a score of about 2 at age 90. People educated 4 years declined from a score of 2.2 to about .9 across the same age range.

So it seems like the real conclusion is that from this data is that no matter what your education level, there's no point in living past 110, 'cause you won't have any cognitive function left.

I, for one, plan to start driving really fast when I hit 70.

John

Posted by: JohnS03 at January 9, 2007 11:47 PM

On the assumption that our learning and memory are products of an associative memory built on a neural network architecture, would we not expect overloading to occur in cases where a greater volume of information has been stored? (This runs on my hypothesis that senility in people is analogous to overloading in neural networks).

Best wishes!

Posted by: Chris at January 10, 2007 4:14 AM

Excellent stuff all around here -- thanks for the clarifications on what's actually in the paper.

Chris, I love the idea that senility is about overloading!

John, an excellent plan.

Posted by: Clive at January 10, 2007 2:09 PM