Slashdot Mirror
UK Council To Send Obese People 'Motivational' Texts Telling Them To Use Stairs
Qedward writes "Stoke-on-Trent City Council is sending texts to obese people in the area to help motivate them to lose weight. Examples of the texts sent include 'aim to eat a variety of fruit and vegetables each day,' 'aim to eat regular meals and keep a check on snacks and drinks' and 'maybe walk to the shops or use the stairs more often.' Over 100,000 people in the region are overweight or obese, the council said, and the texts are for those who are aged at least 18, have a body mass index of 25 or over and who have proactively signed up to receive the motivational messages."
But then I read the last line of TFS.
This is OPT-IN. You only get them if you sign up for them.
Of course, at some point, they will tap into NHS (I'm assuming that this is UK), and send to everyone, regardless of whether they had opted-in or not.
General Relativity: Space-time tells matter where to go; Matter tells space-time what shape to be.
I don't think it's a surprised that people in Cambridge, for example, are quite healthy and fit and people in Stoke are not. It's not just a case of wealth but people in Cambridge are better educated and probably feel they have more to live for. The UK government has drained a lot of the country's ability to compete in favour of advancing London. It's no surprise people in poorer areas probably simply don't care. What's the incentive to lead a healthy life when you live in one of the more economically depressed areas of the country and the government makes it clear you don't matter because you're not a banker?
"Get off your arse and walk, fatso!"
"Put down the fork, you slob!"
"You're disgusting! No one will ever love you!"
"Be sure to vote in Council elections next Tuesday!"
Actually, if you RTFS, they signed up to receive the messages. So it's called for.
True. BMI is hardly perfect... but it's a good first step. Kinda like "Did you reboot it?"
Actually, no it's NOT a "good first step." It's a fair to poor "first step." It's more like a troubleshooting procedure that's guaranteed to give bogus results in a significant percentage of cases -- i.e., where it's wrong or off so often that asking the question is almost more likely to lead you down the wrong path for many cases, rather than giving you useful information.
The BMI formula assumes that adiposity relates to height squared. It does not. This is a simple fact. When people get taller, their bodies scale in three dimensions, not two. So, for really tall people, it's guaranteed to say they are more fat than they really are, and for really short people, it's guaranteed to say people are at a healthy weight when their fat or even obese.
The only reason the BMI formula appears to work at all is because women are both shorter and tend to have higher "healthy" bodyfat percentages. The BMI ranges are more-or-less supposed to be the same for men and women -- that should be a major red flag to anyone who knows anything about bodyfat, since healthy bodyfat ranges for men and women are clearly known to be different.
So, the mean female height is less than the mean male height, but the mean healthy female bodyfat amount is higher. Thus, to have a formula that works for both sexes, you need something that doesn't accurately reflect a "normal" body being scaled up or down. BMI fits the bill, and thus it has been used for population studies to compare mean statistics for overall populations. For individuals -- which it was never designed for -- it's TERRIBLE.
You can immediately see that from the men vs. women problem. BMI says a short man who is in the middle of the height range for women should have the same weight as an average woman. Given what we know about bodyfat, that doesn't make sense. Or, a tall woman who's over 6' or something -- to have a "healthy" BMI she'd often have to look like a waifish model.
"But," you say, "it's still a good first step. It's a simple formula."
Nope -- doesn't excuse it. There are a number of studies that have shown that a number of even simpler measures are actually more accurate at predicting health problems, propensity for disease, etc. For example, simply measuring the waist circumference for men -- regardless of height -- has been shown to be a better predictor of health problems than BMI. Think about that for a moment. Clearly a guy who is 5' tall should not have a waist size the same as a 7' tall guy. But studies have shown that even measuring the waist and saying, "Is it bigger than X inches?" without knowing anything about height, weight, or anything else is a better predictor than BMI.
Yeah, BMI sucks that bad.
But could a simple formula do that much harm? Well, why not just modify that "simple formula" to make it more accurate? In the days when you had to calculations by hand or with a slide rule, a formula involving only an exponent of 2 and a division might make sense. But most people don't calculate BMI by hand anymore -- they plug things into some sort of web calculator or look at a chart.
We can easily fix BMI to make it much more accurate. First, just change the exponent. Logically, as I mentioned at the outside, squaring the height makes NO SENSE. You might think that cubing the height would be better, since the body expands in three dimensions, but it turns out that the male/female factor and other things that don't quite scale precisely with the cube of the height makes an exponent of 3 bad too.
Various empirical studies have suggested an exponent of somewhere in the 2.3 to 2.7 range would be better. But really, to get any accuracy at all, you'd have to at least consider separating the sexes. At that point, you could narrow the range of the exponent for males an