Obesity Map
This obesity map of the United States is currently under discussion at The Usual Suspects. The interesting thing is that, according to the map’s data, obesity in America has sharply increased since 2004. As Beth points out: “What in the hell is going on? That’s not just fast food and television; those things have been around for years.”
Some interesting comments from the thread:
A few months ago the meme that fat may be contagious was all the rage. If that hypothesis is correct it could just be that we reached a tipping point in fat acceptance. My own body shows you can gain a lot of weight quickly if you’re not eating healthy or execsising for a year.
I think people are just starting to be honest. I’ve noticed friends of mine who are now starting to tell the truth in their online personal ads, for example. I admit my weight a lot more easily than I used to and I’m the largest I’ve ever been (244 or so). It seems easier now since other people are being honest.
BMI has some pretty severe limitations, but it’s pretty ominous that it’s rising so steeply. (It is not very likely that we are abruptly getting shorter. It is possible that we’re getting more truthful about our height and weight)…I look at changes that steep in usually slow-changing items and typically suspect changes in survey methods or demographics, but I haven’t been able to find any big obvious changes. Then again, I’m no expert on obesity or national health surveys.
One point this report makes is that, because BRFSS can only call land-line numbers, the increasing use of cell phones might be a pretty substantial confounder. If only poor (or honest) people answer land lines, and cell phones are more common now than they used to be, there will be changes in telephone survey results even in the absence of changes in the population’s health.
So what do you think. Is this really “ominous”? Is the data being manipulated or is it flawed in some way? If the statistics are legit, what happened between 2004 and 2006? And does it matter?
Posted by mo pie
Well, there is one flaw, for better or worse, it is based on self-report phone interviews. This means that people are not being weighed by a consistent and accurate scale. As for the rest, you’d have to pull the research to determine if the information has other weaknesses.
I think the short answer here is that we really don’t know what’s going on, and there’s no way we could know based on this single study. All of the possible explanations postulated over at TUS sound like worthy subjects for followup studies.
The finding that BMI has been going up rapidly over the last 20 years or so has been corroborated by several studies using alternate methodologies, so I think that’s probably a pretty sound finding. If the increase from 2004-2006 shown in this study is real, it will also be corroborated by future studies.
My question is, is the average BMI going up, -OR- are more people “obese” because they changed the definition in 1999?
Average BMI is going up. The pre-1998 classifications didn’t even have an “obese” category (they called it “severe overweight”) so when you see these charts, they reflect the modern classifications even though they partly use old data.
OK La Wade, there is an increase. We get it. So, what would you like to do about it? I always enjoy the condescending tone of your posts, by the way.
Sorry, couldn’t resist.
First of all: why the inconsistent jumping of years in the CNN slideshow? For instance, from 1985 to 1991 takes 6 years, and from 1991 to 1995 takes 4. In all the slides combined we have intervals of 6, 4, 3, 2 years. Whereas the source, the CDC, had surveys done every year. So why the jumping, which I believe is not statistically sound? My guess: for dramatic effect.
But second, and much more importantly: the CNN map is INCORRECT. If you compare the 2006 map on CNN with the CDC map of 2006 (see http://cdc.gov/nccdphp/dnpa/obesity/trend/maps/index.htm) you’ll notice differences where there shouldn’t be any, as CNN states they used the CDC data!
Of course, the CNN map turns every state with a rate above 25 percent into a deep red, where the CDC has two categories, 20-25 percent and above 30 (and uses orange and deep orange, instead of red). This makes things look “worse”.
But more intruigingly, the centre of the country – from ND to TX (I’m not American, I know most states, but to be on the safe side I’ll use the abbreviations, so you can compare easier) – on the CNN map is completely red. So above 25 percent obesity. This is not correct according to the CDC map. ND, SD, MN and IL should be DARK BLUE in the CNN version, as they are yellow (20-24 percent) in the CDC map. This also goes for WA, OR, ID, NV in the west and the smaller states MD and VA in the east.
HI, MA and RI are dark blue in the CNN map (20-24 percent), but should be as blue as CO (so lighter), as their obesity rate is 15-19 percent, not 20-24 percent like CNN wants us to believe (?)
So yes, it’s getting much and much worse, if you’re not stating the data as they should be. We’re not even talking about manipulating data here (lies, damned lies and statistics), it’s just plain false representation. I can’t understand why though. CNN states its source as BRFSS, CDC. Did the graphics editor who had to change the colors – to make them look more like CNN’s instead of CDC’s – make mistakes, was he tired, colour blind or just plain stupid? (I don’t even want to go into conspiracy theory thinking …)
But this is, of course, how hysteria forms: on the basis of wrong “facts”. But maybe that could be a good thing, fat accepting people could – once again – say the data don’t add up. We’re being lied to – in this case by CNN – whether that’s been consciously done or not.
PS: do correct me instantly if I’m wrong, I’m still a little baffled (and unbelieving) about what I just wrote – maybe someone has a sound and simple explanation why the CNN map is so different from the CDC’s?
I would imagine the data is definitely flawed because of the only calling home phones thing. but, I’m not a professional surveyor, so what the heck do I know?!?
Dutchy, you’re right that the 2006 CDC map is different from the CNN one, which is weird because the 1985-2004 maps are the same as the CDC’s. You’re also right that they skip certain years, but I don’t find that to be a distortion, because if they had actually showed evenly spaced years, it would make the change look more dramatic.
I don’t know if the disparity in the 2006 data is a mistake or perhaps one of the maps is based on more complete data than the other. At any rate, I don’t think it negates the overall finding of big changes in the obesity rate over the last twenty years.
On the land-line phones thing, I have another idea: Most people I know who are my age (late 20s, early 30s) don’t have land lines at all. My guess is that their sample skews older, and therefore heavier. But that’s totally a guess.
Nice catch, Dutchy.
My son doesn’t have a land line.
If I get called for a survey I refuse to participate, if I even answer it. I usually let the machine pick up. I ask how much it pays and when they tell me I won’t get paid, I repeat ‘Do Not Call List’ until they hang up on me.
I would like to see these numbers graphed alongside statistics on percentages of racialized persons in these regions. BMI measurements, in addition to everything else bad and stupid about them, are also radically eurocentric.