Nick, Chris, Fraser and Peter discuss the purpose of BMI

For more information on Aleph Insights visit our website https://alephinsights.com or to get in touch about our podcast email [email protected]

Hello and welcome to the Cognitive Engineering podcast produced by Tell Me Studios for Aleph Insights. In this series of podcasts we take a look at interesting topics and discuss what we think they tell us about analysis and decision making. I'm Fraser McGruer and I'm here with Peter Coghill, Chris Wragg and Nick Hare of Aleph Insights and this week we're discussing Nick Hare's weight.

So Nick, tell us about your weight. Well, it's in my view a little bit too high. So like I have done a few times in the past, I'm currently on a diet. Managed to lose about three kilos since the beginning of the year, which is good. That's on track. It's a sort of year long diet. And I do it very, as you'd expect, very sort of scientifically as it's all based on calories in versus calories out. I weigh myself every day. I take a sort of rolling average, work out how many calories I'm allowed and you know, just make sure that I'm keeping within that. And then, you know, inevitably the weight goes down. I'm only interested in the weight in kilos. That's the only thing I'm looking at. It's the most measurable thing. It corresponds to a, you know, a real physical quantity. So it's, you know, it's a very handy thing to keep track of. But my scales are quite sophisticated and they give me other data as well, which I tend to, I tend to ignore. But I noticed that it had my body mass index. It works that out because I've told it how tall I am. And apparently it's about 30 and a half, right? Which, if you, if you look at the, if you look at the, you know, the kind of labels that they put on these things, 30 and a half corresponds to obese, which is, I don't know, I've, I've, okay, it's arguable. But I was quite surprised. I'm happy that I'm sort of overweight. I will definitely agree with. But I don't know, I kind of, when I think of an obese person, I think of someone who's a little bit bigger. And, and I just sort of got me thinking, why do we have the BMI, right? Because it's just calculated from your weight. And your height. In fact, it's the square of your height. It's not taking into account all the muscle that you've got. All the massive beefy muscle. But also, I mean, it's, first of all, I mean, you can criticise the, the body mass index for a number of reasons. First of all, it's not an index. It's got units. It's in units of kilograms per metre squared. But, but put that aside, it's the fact that there's also the issue that it's, it's calculating, for some reason it uses the square of your height, which is mad. Because if you imagine shrinking me down to half size, my mass would fall by an eighth. But the square of my height would only fall by a quarter. So my BMI would, would, you know, suddenly go, go down massively. So, I mean, one of the reasons, so, so there's, there's a, there's a problem. There's kind of methodological problems. The key thing is, why is it there at all? What's the point? If you already know someone's height, and you know their weight, calculating their BMI is not giving you any extra information. So apart from sort of spurring me on a bit and thinking, oh, flipping heck, I'm obese, I really had better stick to this. I can't, I just, I just wonder what the point of it and other indices like it are.

Okay, well, first of all, you've, you've asked a question there, what's the point of it? So we'll answer that question. But maybe you've answered the question, because you said that then spurred me on to do something. So there you go, end of podcast, you answered your question, it fulfilled a really useful function. But hold on. I know Chris wants to come in. But just before you do, let's just do a quick check on this, because I think the BMI over obese is at 30, where it kicks in for obese. Is that right? Supposedly. Okay, let's just do, out of interest, let's do ourselves. So you're about what, 6'2", something like that? No, 6 foot and a crucial half an inch. Okay, I thought you were taller than that. Anyway. No, I seem taller, because I have such presence. Yeah, exactly. It's your bearing. So Chris, your height, your weight?

Same kind of height. Yeah. 183 centimetres.

Okay. And what's your, how much do you weigh in kilos?

68.

Okay, wow. Okay. So you're probably, you're probably underweight on the BMI, I would have thought. No?

No, no, no. Quite a long way. You've got to get quite low in weight to be classified as underweight.

It's hard to imagine. What would it look like if you were significantly thinner than Chris? You'd be skeletal.

Yeah, true. That's the point. I mean. Well, hold on. We'll come to you. Yeah. And Peter?

I'm exactly 183 centimetres as well.

Is that also six foot and a half an inch or something? It's actually company policy. Yeah, you must be. It's like the guards or something.

Yeah, there's a very strong correlation between being allowed in Aleph Insights. What's your weight? It's 94 kilos on last weighing. And that puts me, that hovers me about the obese normal range. And that's where I've always been. Even when, even a few years ago when I was cycling a lot and being very fit, I was always hovering around there. And I think that touches one of the problems. It doesn't, the BMI, one of the failings of BMI is it doesn't deal with athletic people very well. Not that I'm hugely athletic.

Okay. No, no. Well, you're terribly athletic. So we'll come, we'll hold that thought. We'll come back to that. Hey, we haven't had me yet. We haven't had me. So I'm, I'm, I think I'm exactly six foot, maybe just a shade under, but I'm exactly six foot. A big short arse then. And then, and then I'm, I've lost about 10 kilos last year. So I'm 85 kilos at the moment. And I want to, but that still puts me overweight actually, 85 kilos, but I'd like to get to 80. But anyway, I think just actually, just as an aside, what's interesting, I know my, I never do my height in centimetres. It's always in feet, but I always do my weight in kilos, but that's another conversation.

Chris. Yeah, no, I just, I just wanted to not necessarily speak in defence of the BMI, but explain why I, why I think it's become an instrument of public health policy. And yeah, just sort of talk a little bit about, you know, why, why it might be an effective thing. I mean, first of all, like all things that are any good, it was invented in the 19th century. So that's, that's the first point. Secondly, I think, yes, so there are issues with it, but essentially what it's trying to do is estimate your body fat percentage. And in a general population, it's a pretty good way of doing that. So it doesn't take account of people with a higher muscle mass percentage than the general population. So you've got Olympic athletes who are classified obese, but they've got very low percentage of body, body fat. But in a general population, it's a pretty good estimate of that. And, you know, it's, it's a ready way of saying, okay, for somebody's height, kind of, you know, you, you can straight away start thinking about a series of boxes you might put them in, in relation to their, their body fat percentage. It's, it's a very, so you could measure it more accurately. You could use skinfold calipers is one measure of measuring body fat percentage. You could look at hydrostatic weighing where you, you weigh somebody in and out of water and you calculate their, their body density and all those kinds of things. So there are other better ways of, of doing it, scanning technologies as well. But for simple cheapness, weighing somebody and measuring their height is, is pretty good. And then coming up with some determination about them because it's been around so long, there is lots of data built on these categories of, I think it's, if your, your body mass index is below 18.5, you're considered underweight. If you are 18.5 to 25, it is, you are considered to be in a normal range, over 25, but lower than 30 is overweight and above 30 is considered obese. And then there are other categorizations within those such as, you know, morbidly obese, et cetera. But the point is from public health policy perspective, they're looking to essentially say, do we need to take any action about this individual? Is this person perhaps subject to eating disorders? Do we need to, you know, have some intervention there? Is this person in a sort of warning area where we can arrest something and, and bring them, bring them back again? Or are they in a stage where we perhaps need to take more dynamic action? And for, for an industrialized healthcare system, having four boxes is quite, is quite useful. And having an easy way to put people in four boxes is, is quite useful. And we can all, I think, accept that the overall aim of, of looking at body fat percentage and trying to reduce body fat in an individual is a health, a valid healthcare goal.

Great. I don't think you said anything that I disagree with. And I think it answers Nick's question. As I say, indeed, I think he answered himself. What's the point of it? The point is it spurred him into action. But Peter, you were saying something about athleticism. I mean, have you, are you, you were going to say something about athleticism and not taking anything into account? Or is there anything that Chris said that, that, that doesn't tie in with what you're about to say?

Not, well, apart from those, apart from the exceptional cases, people of particularly small or tall stature, people of particularly dense muscle build, the athletes and things. But they, but I think Chris's point is correct. And the bands that define whether or not you're obese, beast, normal, underweight, etc. They are, they're there because they, those bands correspond well with illnesses that have strong correlation with a particular type of body. And it is a crude measure, but it seems to work for most things. So as a GP, if you present with symptoms, and they can take your BMI, then it tells them quite a great deal because it will put you into particular classes of people that make you much more or less likely to have certain illnesses. And Chris's point again, it's a simple measure. It's very, very easy. And you can do it at home with a set of scales and a ruler. You don't need any specialised equipment. It's easy to apply.

Okay, so that being the case, end of podcast, no? I mean, what I would say, I want to come to Nick. So either, what I would say to you is, is either, have you got an alternative measure that you can come up with that's as simple and as easy to use? Or tell me, talk to me about something else if there's something that's more interesting than that question?

Well, actually, if they just use the cube of height, that'd be great. I think that would just, it would satisfy me because it would make sense. But no, I'm actually, I'm fine with simple indices. And I think Chris touched on the key point here, which is, how do you know if an index is any use? It has to correlate with something real, right? That's the key thing. It has to correlate sort of in a, you know, empirically with the thing that we've supposed you can't measure, right? But which we, which we're really interested in. So in this case, you know, we think that BMI might, let's propose the hypothesis that BMI is correlated meaningfully with, you know, with things like type two diabetes, or coronary heart disease or something. And, you know, and then you don't have to, you know, do expensive tests to try and measure people's, you know, measure people's arterial, you know, walls and that kind of thing, because you because you have this thing that we've sort of proved through studies, actually, it's just correlated with that anyway. But I suppose that I think what happened, what the problem is, when people get attached to the indicator, you know, when people get attached to the thing, which is actually only a proxy in the first place. And, and then, you know, and then and particularly if people start sort of targeting it, you know, if people so so, you know, if I know, most athletes are aware of this issue, but you can imagine someone who is actually very fit, looking at discovering that they're clinically obese, and sort of taking action on that, despite the fact that it's actually only supposed to be an indicator of something, and we know it doesn't capture. The thing we're really interested in, which is your, let's say, your primarily your body fat, there are there are other indices that have been proposed, like, I mean, apparently, your waist measurement is, is, is better, is more diagnostic, actually, and it's kind of easier to measure as well. But, but, but yeah, I mean, I mean, that's, that's the issue, really, it's sort of whether or not people start use start hanging off the index, and instead of that, instead of actually, you know, let's say we can measure body fat, if we can measure body fat accurately, which, you know, my scales do it using my skin conductivity or something, then we don't need to worry about BMI, we don't, but but you can imagine someone even with low body fat still thinking, oh, my BMI is too high. So, you know, that that's the problem is if we're not aware of its limitations. So let's not get hung up on the proxy.

Chris.

Yeah, no, I was just gonna say, I mean, the waist measurement one's a really interesting one, because, you know, more than your overall percentage of body fat, the way that adipose tissue is distributed is more important, and particularly around the the abdomen, you know, that's that's associated with increased risk of disease. And, you know, I can see circumstances where, given the way technology is going, and given the lack of cost associated with processing complicated data, that you we might actually move beyond the BMI, you know, the BMI is something relatively easy to, well, very easy to record now. But actually, there are other measures that are going to be equally as easy to make in the future, or combinations of measures. So it might be it's very easy to get somebody's BMI to measure their waist to conduct some other measure of body fat percentage or some other risk factor, and jumble them all together and come up with a composite factor that is a better predictor. And to use that, but for the time being, BMI kind of probably does its does its job.

So it sounded like one of the things you were saying there is potentially there's a technological solution or solutions, which can be better proxies. And fortunately, we've got a technological expert with us. And Peter, you were nodding your head saying you were you were you were nodding your head vigorously at one point when Chris was speaking there.

Yeah, I think well, the point the fundamental point is that all these, so these better alternatives to BMI rely on more data. So or harder to measure data. So there's more cost involved. But yes, as if once we're in a world where everyone is wearing a Fitbit, and we're able to harvest data about how much we move around from our phones. And we just much more instrumented, we'll be able to build much more complicated models of health, which will provide us with as simple to apply proxies for overall healthiness as BMI. And when presented when presenting with symptoms, that your first line health, first line health care will be able to put you into a much finer bucket, because they'll have historical data about the population as a whole, your type of your class within the population, but also specifically you as an individual.

Because, because, you know, that that's a good point, because ethnicity is a factor as well, actually, that in certain ethnic groups, the most of the data around BMI has been done in sort of, you know, Western European, North American populations, or one skewed towards particular ethnicities. And they've actually found that the disease incidence in other ethnic groups doesn't correspond with the same thresholds that you might, you might have a lower BMI, but may be more susceptible to particular conditions if you come from a particular ethnic group. So that's something straight away, you could incorporate into your into your data set and start, start using.

I mean, we've got this all sewn up, right? I mean, there's I mean, what more, what more is there to say here?

Well, I think there. So I think the the BMI itself is crude, we acknowledge it's crude. But to say it is crude, it's been pretty effective. It's been a pretty effective tool for public health. But, but a fairly blunt one. And I think, you know, we're now moving towards a stage where more individualised healthcare is possible, you know, much more is known about individual genetic markers of disease that you can start to start to measure. But also, you know, with all these wearable technologies, we're getting more and more activity data about individuals. And, you know, you can potentially measure somebody's food intake much more more accurately. And I suppose it's it's implausible for a healthcare professional to look at every individual as an individual and to look at all of that data. But it is not implausible for, for machines to look at every individual as an individual and start to prescribe on a more individual basis. And I think that's where we need to start, you know, investigating alternative measures and, and then testing their their predicted performance of disease.

Okay, Peter, sorry.

I didn't know this until last night, but in France, Italy and Spain, you can't, you're not allowed, you can't be a professional model on a catwalk, if you have a BMI less than 18. Which, which, yeah, okay, so there's obviously well intentioned drive to try to make models more healthy, not be so not be so unhealthily skinny. But what happens if you've got a seven foot model? Their BMI will, because of this, because of the square problem... Contact the nearest circus.

Actually, it's the short, I think it's short, short models who would suffer. Their BMI would because you're dividing by, yeah, dividing by a bigger number.

So if I'm lobbying on behalf of models of short stature, then, then I'm going to meet this, this seemingly arbitrary piece of legislation, which prevents my, my clients from, from, from taking part in the profession that they've chosen. Yeah, yeah, yeah.

I mean, I think it's the, there's a lot of the way that, I mean, you know, I think a lot of us don't realize how much the, the advice that we hear and potentially follow is predicated on a lack of information. I mean, so take something really obvious, like telling people to give up smoking, that, you know, the, the only reason we do that is because we can't tell whether someone is going to get, you know, bronchitis or lung cancer in time to, to do something about it. Now, if you can imagine, imagine sort of a cancer detector that was able to tell you the minute you had some, you know, the sort of first sign of cancer and, and there was a very sort of easy, relatively non-invasive invasion to remove it. In that world, we could just get do away with, you know, don't smoke. It's, it ceases to be an issue. The only reason we tell people not to smoke is because we know that it's highly correlated with these diseases and we cannot cure them in time to stop you dying. And actually that, that's a, you know, just one example, a lot of medical advice, a lot of advice about lifestyle in general is based on the premise that we don't have enough information to be able to, to, you know, to actually take account of your individual case.

Well, I think that's an admirable aim and I applaud that. I also would like there to be some technology developed that would allow me to smoke because I really enjoy it, but I can't do it.

Well, it's interesting. I mean, that's sort of, I mean, this is slightly related to the, to this issue of people getting hung up on the, the index or the thing rather than the thing we actually care about is the, the sort of quite vehement lobbying against electronic cigarettes, which Peter and I are both smokers of, or I should say vapors of. It's almost like people, you know, we have discovered that smoking is bad for you. So almost as a kind of, in a, in a sort of like a cargo cult, it's like actually things that look like smoking must be bad for you as well, you know, and sort of pubs are now increasingly banning the use of electronic cigarettes, even though, well, I mean, you know, and I suspect that would be happening even if you could prove that there was no danger from secondhand vapor. Okay.

Unless anyone's got anything to say, we need to wrap up there. Okay. Right. No one's, no one's jumping in. So that's good. Nick, quick question. So do you have an aim in you? Because I mean, you've got the, despite, despite being a bit dismissive about the BMI index, this has already resulted in you, this index has resulted in you losing three kilos. No, no, it's nothing to do with that. No, it is. No, it's not.

It's looking, looking at myself in the mirror and going, flipping heck, I'm fat. That's the, that's the big thing. No, it's the, the, I'm only looking at weight. So I get, I've got a target. Fine. What's your target? 83 and a bit. So my, my, my target where I'd really like to get to is, is massively higher than Chris's actual weight, which is, you know, interesting. I would, I don't think, I don't know how.

Yeah, and I'm on a similar diet as well. And I have, since about 18 years old, never been lighter than 82 kilos. And I get to 82 and no matter what I do, I can't get any lower. So that's my sort of, my realistic ceiling.

Because I too have, I'm more or less constantly on some sort of diet or something or some exercise kick. And, and I, my, my, my, my long-term goal has always been 80 kilos, 80.0 kilos. And it has been that for about eight years.

We're kind of clearly 80, 80 kilo guys. But you look at Chris, he doesn't look like he, I can't imagine him being, I mean, we're kind of just naturally clearly bigger looking.

How bloated would Chris look at 80 kilos? This is what I'm wondering. We need to do that. We need to experiment on Chris.

He's thinnest. We couldn't ever hope to be as sort of thin as Chris, but I don't want people to go around thinking he looks like a kind of, you know, a sort of Biafran refugee. We do feed him. He does look like a normal human being, but it is extraordinary that. Certainly in terms of weight he does. It goes to show what, you know, the variation you can get in, in sort of what normal is.

So just to sort of finish off that point, given that there are clearly two different types of people in this room, people who are around 80, 90 kilos and don't look too bad for it. And then people who are around 60 kilos also don't look too bad for it. This is important. There are different classes of people that are still healthy enough within their own normal band. So your BMI should be sort of, should be more, if anything, more intelligent and take that into account.

Okay. The most important thing is that you should be you.

And just be happy with yourself.

But actually what I would say is actually that of the two classes of people in the room that we talked about and we said, actually, they both look okay. I disagree, actually. I think one of the classes of people of which I am one don't look okay at all. But anyway, I just want to look like you, Chris. That's all I'm saying. Okay. So we'll wrap up there. Thank you for listening to the Cognitive Engineering Podcast. I'm Fraser McGruer. We've been here with Chris Wragg, Peter Coghill and Nick Hare of Aleph Insights. And until next time. Thank you. Bye-bye.