Join 3,430 readers in helping fund MetaFilter (Hide)


Insert You Momma So Fat Joke Here
January 10, 2007 7:46 AM   Subscribe

High BMI Now Means Cognitive Difficulties Later? A study published in Neurology attempts to discover if there is a link between cognitive function, cognitive decline and BMI (body mass index) over time. Yes, I am aware that BMI is a flawed metric.
Full Text (sub. req'd).
posted by fenriq (32 comments total)

 
Erm.... could the link be more lack of proper nutrition > obesity > cognitive problems in old age?
posted by jokeefe at 8:29 AM on January 10, 2007


Hang on, are you saying fat people are stupid, or stupid people are fat, or fat people progressively become stupider, or stupid people progressively become fatter?

/me == fat and stupid, and getting fatter and stupider by the minute
posted by unSane at 8:40 AM on January 10, 2007


We haven't had a good fight over physical attributes for, oh I don't know...ages.

May I be the first to call for Wendell on this one.
posted by Sk4n at 8:54 AM on January 10, 2007


Hang on, are you saying fat people are stupid, or stupid people are fat, or fat people progressively become stupider, or stupid people progressively become fatter?

Nay, simply that there's a correlation between a higher BMI and cognitive tests. They reason and validity for the correlation is up for grabs.
posted by jmd82 at 8:57 AM on January 10, 2007


And by "physical attributes" i'm referring to the male cut/uncut derailing comments that besmirched that interesting post.
posted by Sk4n at 8:59 AM on January 10, 2007


I'm sorry, I don't quite follow your post.
posted by mikeh at 9:09 AM on January 10, 2007


I had an insight the other day. Certainly not original; only I never thought of it before.

If you take a population with lots of different binary (not necessary, but for the sake of the example it'll do) properties, each of which is distributed randomly and independently about the population,

then - despite the independence - there is, for any inferential correlation/causality test, a nonzero probability that some two of the variables will appear to be correlated at the level of significance specified by the test (even though they really aren't).

This probability increases (I haven't computed how fast) as the number of properties increases.

The moral: if you study enough pairs of properties in people, you will eventually find stastical evidence that some pair is correlated, even if none of them are.
posted by Wolfdog at 9:12 AM on January 10, 2007 [1 favorite]


Obese patients seem to recover from acute heart failure better than any other group. Now who's stupid?
posted by landis at 9:16 AM on January 10, 2007


Does this mean that Santa gives me things because he's dumb?
posted by Pecinpah at 9:18 AM on January 10, 2007


Some of the information is logically sound...higher percentages of body fat could also mean higher cholesterol, which would clog blood vessels and reduce blood flow to the brain.

But really, I just came in to say ya mama's so fat, she got hit by a bus and said, "Who threw that rock?"
posted by deusdiabolus at 9:26 AM on January 10, 2007


This data can be boiled down to the more obvious: physical inactivity/depreciation may cause certain forms of atrophy. It should be noted that low weight (be it from malnutrition, self-abuse, illness, or birth defect), in severe cases can lead to dementia or may contribute to schizophrenia in offspring.
posted by Smart Dalek at 9:27 AM on January 10, 2007


I'm curious for somebody with access to the full article:
Do they actually go into detail about why people have higher BMI? Does x person with z BMI exercise, eat haelthy while person y also has z BMI but does nothing healthy?
posted by jmd82 at 9:44 AM on January 10, 2007


I love that we've gotten to the point where it's necessary preemptively to disavow the BMI before referencing any studies that involve it.

A p-value of .001 is pretty weighty (har har). Just a few thousand more studies like this, all pointing out the staggering health consequences of the obesity epidemic, and maybe things'll start to get done.
posted by gurple at 9:45 AM on January 10, 2007


"I've read that same formula, but as an athlete, I'm classified as phenomenal," O'Neal told The AP. "You can look it up."

Shaq just slays me.
posted by Kwantsar at 10:29 AM on January 10, 2007


deusdiabolus, that was the conclusion of an article I read about this offline.

gurple, I'm waiting for someone to revise BMI to incorporate info about the hip-waist ratio data to make it more valid again. Alas, I'm still waiting.

By the way, I'm not "saying" anything, I'm merely the conduit for the researchers to say something. They are saying that a higher BMI now means that you are statistically more likely to score lower on cognitive function testing down the road.

As deusdiabolus notes, it may be partially because of physical obstruction of the nutrient pathways in the body.
posted by fenriq at 10:35 AM on January 10, 2007


As we all know, there are two kinds of people with high BMIs: fat people and muscular, athletic people. Obviously, the dumb jocks are dragging down the average for us fatties.
posted by Faint of Butt at 10:43 AM on January 10, 2007


What other variables did they measure? Seems to me that this, if a valid relationship, could be associated with a number of things that affect BMI, including stress, hormonal changes, diet, and medication.
posted by zennie at 10:50 AM on January 10, 2007


Gurple, it's my understanding (IANAStatistician) that a p-value of .01 is just as weighty as a p-value of .001 in that they both are arbitrary limits decided upon by the researcher. The more cases (observations) you have, the more likely you'll see smaller p-values; but smaller p-values won't increase the probability that the null hypthosis is true.
posted by parilous at 10:55 AM on January 10, 2007


What other variables did they measure?

Penis Size. It seems these dumb fatties also have abnormally HUGE winkles.

Yeah. I know. I don't like god's sense of humor either.
posted by tkchrist at 11:06 AM on January 10, 2007


P of 0.01 means that there is a one-in-a-hundred chance that the observed result is the product of chance (ie, just a weird distribution, not reflecting an underlying correlation). P of 0.001 means that the chance is 1 in 1000.

The usual limit for statistical significance is generally P less than or equal to 0.05
posted by Mister_A at 11:11 AM on January 10, 2007


My father has been thin all his life. Late in life, he's having major cognitive difficulties. So either this study is flawed, or my father's rockin' at the extreme.
posted by davejay at 12:33 PM on January 10, 2007


Excercise promotes neuronal growth!
posted by porpoise at 12:33 PM on January 10, 2007


Why, if we know that BMI is a flawed metric, is it okay then to post references to studies that use BMI as if they tell us something?
posted by kalessin at 12:52 PM on January 10, 2007


There's a big difference between flawed and useless. For most purposes, including from what I can glean this study, BMI works just fine.

Anyway, fatties are stupid. Obvious. Next.
posted by wilful at 4:45 PM on January 10, 2007


Okay, well maybe if I repeatedly use flawed statistics to prove some point that's on my agenda (like white people are stupid or like men are stupid), I can get something "proven" here on Metafilter too.
posted by kalessin at 5:18 PM on January 10, 2007


See, flawed BMI does not automatically and exclusively mean flawed study. That's your comprehension problem right there. I have no agenda - are you fat?
posted by wilful at 5:38 PM on January 10, 2007


I don't know, wilful. Are you a white guy?
posted by kalessin at 6:18 PM on January 10, 2007


kalessin, BMI is flawed in that it says people with higher amounts of muscle can be considered obese when that's just not the case. The flaw applies primarily to athletes and isn't especially applicable here. I pre-empted the kvetching with the recognition of its inherent flaws but it did not work.

According to the BMI, I am overweight at a 25.4 (just barely) but I am a fairly muscular guy with the same size waist as when I was in high school twenty years ago.

BMI crossed with fat percentage or against the hip-waist ratio would be less flawed but no system of categorizing huge (pun intended) swaths of the populance is going to spot on. As davejay notes anecdotally above.
posted by fenriq at 6:28 PM on January 10, 2007


define white, and guy, using a non-flawed methodology.
posted by wilful at 6:32 PM on January 10, 2007


Okay, so even though BMI is flawed, it's still okay to call fatties stupid. Gotcha.
posted by kalessin at 6:33 PM on January 10, 2007


wilful, you first, big guy!
posted by kalessin at 6:34 PM on January 10, 2007


who you callin' big?
posted by wilful at 8:23 PM on January 10, 2007


« Older Downtown Lives On...  |  Arcade '84, from Cinemarcade.... Newer »


This thread has been archived and is closed to new comments